Selecting the optimal activation function for training deep neural networks has always been challenging because it significantly impacts the neural network’s performance and training speed. At this point, researchers are more likely to employ RELU than other well-known activation functions. After RELU, researchers have proposed many activation functions to improve RELU. None of them was capable of surpassing RELU as their most significant rival. SWISH outperformed RELU in several challenging experiments like classification, object detection, and tracking. Replacing RELU units with SWISH, which improves performance in many tasks. This paper proposed a new activation function surpassing Google’s brain’s SWISH function, Which we named AIF. Experiments indicate that our proposed activation function outperforms SWISH in various tasks.

In urban areas, deep excavation-induced ground deformation may damage adjacent existing structures and is conventionally evaluated by levelling at installed settlement points. However, a small number of measurements cannot represent the total change in ground deformation adjacent to excavation sites. Furthermore, significant local subsidence may occur in places where settlement points have not been installed and only noticed after an accident. For deep excavation sites located in urban areas, paved pedestrian sidewalks are often located adjacent to sites, and construction activity can cause these paving blocks to displace. This study introduces a method to detect paving block displacement adjacent to deep excavation sites using terrestrial photogrammetry. A digital camera creating point cloud data (PCD) and an acquisition method satisfying the frontal and side overlap requirements were demonstrated. To investigate the displacement detection and measurement capabilities by PCD analysis, an experimental program was conducted, including a PCD comparison containing the uplift, settlement, and horizontal paving block displacement and reference data. The cloud-to-cloud distance computation algorithm was adopted for PCD comparison. Paving block displacement was detected for displacements of 5, 7.5, and 10 mm in the uplift, settlement, and horizontal directions; however, the horizontal displacements were less clear. PCD analysis enabled satisfactory measurements between 0.024 and 0.881 mm for the vertical-displacement cases, but a significant error was observed for the horizontal-displacement cases owing to the cloud-comparison algorithm. The measurement blind spot of limited settlement points was overcome by the proposed method that detected and measured paving block displacement adjacent to excavation sites.

Auto-labeling is one of the main challenges in 3D vehicle detection. Auto-labeled datasets can be used to identify objects in LiDAR data, which is a challenging task due to the large size of the dataset. In this work, we propose a novel methodology to generate new 3D based auto-labeling datasets with a different point of view setup than the one used in the most recognized datasets (KITTI, WAYMO, etc.). The performance of the methodology has been further demonstrated with the development of our own dataset with the auto-generated labels and tested under boundary conditions on a bridge in a fixed position. The proposed methodology is based on the YOLO model trained with the KITTI dataset. From a camera-LiDAR sensory fusion, it is intended to auto-label new datasets while maintaining the consistency of the Ground Truth. The main contribution of this work is a novel methodology to auto-label autonomous driving datasets using YOLO as the main labelling system. The performance of this approach is measured retraining the contrast models of the KITTI benchmark.

Deep learning technologies for skin cancer detection have been dramatically advanced based on high resolution dermoscopic images. The low-cost approach based on clinical skin images of low resolution is promising but remains technically challenging due to their undermined image quality. In this paper, we propose a coarse-to-fine efficient super resolution transformer (CF-ESRT) network to reconstruct the dermoscopy-level high resolution skin image from a low resolution clinical image. By connecting the refinement network to the original super resolution transformers and applying perceptual and gradient losses, our framework noticeably improves the finer texture details of skin lesions in the super resolution (SR) images, and is effective to elevate the perceptual quality of the SR images. Quantitative and qualitative evaluations show that our method outperforms ESRT the basis model as well as the other state-of-the-art SR models.

As AI models become more and more common in business and even in our daily lives, it is important to understand what the carbon impact of these models is. Recent papers have shown that this impact can be quite great, i.e., the training of a single high-end model can result in emissions of more than 500t of CO2eq. In this paper we discuss the factors that influence the carbon footprint of AI models, explore what impact different decisions have, and show how the footprint can be reduced. We also examine the footprint of different models to give a guideline on how urgent action is for different organizations.

One of the most common forms of dementia is Alzheimer’s disease (AD), which leads to progressive mental deterioration. Unfortunately, there is no definitive diagnosis and cure that can stop the condition progressing. The diagnosis is often performed based on the clinical history and neuropsychological data, including magnetic resonance imaging (MRI). Deep neural networks (DNN) algorithms are gaining popularity for medical diagnosis, and have been used widely for the analysis of MRI data. DNNs can extract hidden features from thousands of training images automatically. However, they cannot judge how confident they are about their predictions. To use DNNs in safety-critical applications such as medical diagnosis, uncertainty quantification of DNNs predictions is crucial. For this purpose, Monte Carlo dropout (MCD) has been widely used, however, it may lead to overconfident and miss calibrated results. This paper proposes a framework in which the MCD algorithm’s hyper-parameters are optimized during training using Bayesian optimization for the first time. The conducted optimization leads to assigning high predictive entropy to erroneous predictions and making it possible to recognize risky predictions. The proposed framework is used for AD diagnosis, which has not been done before. We compare our method with some existing methods in the literature based on different uncertainty quantification criteria. The results of comprehensive experiments on the Kaggle dataset using a deep model pre-trained on the ImageNet dataset show that the proposed algorithm can quantify uncertainty much better than the existing methods.

Most machine learning and deep learning algorithms can only use low-dimensional data as input, but the data that must be processed in practical applications is diverse and irregular. There are two main problems with big dynamic data. (1) The size of the embedding table grows linearly with the vocabulary size, resulting in massive memory consumption. (2) For different newly added vocabularies. To solve these two problems, this paper proposes a novel embedding algorithm that can learn attribute associations with entities based on deep and hash algorithms. Taking movie data as an example, the encoding method and the specific flow of the algorithm are presented in detail, and the effect of dynamic reuse of data models is realized. Compared with the four existing embedding algorithms which can fuse entity attribute information, the deep hash embedding algorithm proposed in this paper has obvious optimization of time and space complexity.

Classical religious texts remain an essential part of human culture due to their undiminished influence on the advancement of civilization. Although their entirely divine origin is questioned repeatedly, explicit or implicit quoting and adherence to their basic guidelines are fundamental in modern society. In this respect, these documents’ inner structure and linguistic style appear to be pivotal. This paper considers the topic from the standpoint of small textual patterns classified using deep learning methods, traditionally applied to analyze short textual material like tweets. We divide the considered documents into small sequential chunks imitating tweets and categorizing them, classifying an entire text. The proposed method demonstrates that the religious text collections correspond to stable ”Twitter”-like structures that adequately reflect stylistic properties. So, concise word combinations seem to be an inborn textual attribute that adequately outlines the proposed multi-source authorship. This approach differs from traditional methods of analyzing classical religious documents, which are based on the consideration and interpretation of relatively long templates. The case study consists of three famous collections of Mosaic authorship in the Old Testament (Hebrew), Pauline authorship in the New Testament (Greek), and Al-Ghazali authorship (Arabic). The obtained results go well with most previously expressed evaluations and complement them with new implications, particularly in the authorship of two famous manuscripts attributed to Al-Ghazali.

Convolutional neural network architectures have become increasingly complex, which has improved the performance slowly on well-known benchmark datasets in the recent years. In this research, we have analyzed the true need for such complexity. We have introduced G-Net light, a lightweight modified GoogleNet with improved filter count per layer to reduce feature overlaps and complexity. Additionally, by limiting the amount of pooling layers in the proposed architecture, we have exploited the skip connections to minimize the spatial information loss. The investigations on the proposed architecture are evaluated on three retinal vessel segmentation publicly available datasets. The proposed G-Net light outperforms other vessel segmentation architectures by reducing the number of trainable parameters..

Globally, zoonotic diseases have been on the rise in recent years. Predictive modelling approaches have been successfully used in the literature to identify the underlying causes of these zoonotic diseases. We examine the latest research in the field of predictive modeling that verifies the growth of zoonotic pathogens and assesses the factors associated with their spread. The results of our survey indicate that popular mathematical models can successfully be used in modeling the growth rate of these pathogens under varying storage temperatures. Additionally, some of them are used for the assessment of the inactivation of these pathogens based on various conditions. Based on the results of our study, machine learning models and deep learning are commonly used to detect pathogens within food items and to predict the factors associated with the presence of the pathogens.

Convolutional Neural Networks (CNNs) have gained prominence in the research literature on image classification over the last decade. One shortcoming of CNNs, however, is their lack of generalizability and tendency to overfit when presented with small training sets. Augmentation directly confronts this problem by generating new data points providing additional information. In this paper, we investigate the performance of more than ten different sets of data augmentation methods, with two novel approaches proposed here: one based on the Discrete Wavelet Transform and the other on the Constant-Q Gabor transform. Pretrained ResNet50 networks are finetuned on each augmentation method. Combinations of these networks are evaluated and compared across three benchmark data sets of images representing diverse problems and collected by instruments that capture information at different scales: a virus data set, a bark data set, and a LIGO glitches data set. Experiments demonstrate the superiority of this approach. The best ensemble proposed in this work achieves state-of-the-art performance across all three data sets. This result shows that varying data augmentation is a feasible way for building an ensemble of classifiers for image classification (code available at https://github.com/LorisNanni).

The study aimed to identify different molds that grow on various food surfaces. As a result, we conducted a case study for the detection of mold on food surfaces based on the “you only look once (YOLO) v5” principle. In this context, a dataset of 2050 food images with mold growing on their surfaces was created. The dataset was trained using the pre-trained YOLOv5 algorithm. In comparison to YOLOv3 and YOLOv4, this current YOLOv5 model had better precision, recall, and average precision (AP), which were 98.10%, 100%, and 99.60%, respectively. The YOLOv5 algorithm was used for the first time in this study to detect mold on food surfaces. In conclusion, the proposed model successfully recognizes any kind of mold present on the food surface. Using YOLOv5, we are currently conducting research to identify the specific species of the detected mold.

The medial septum (MS), as part of the basal forebrain, supports many physiological functions, from sensorimotor integration to cognition. With often reciprocal connections with a broad set of peers at all major divisions of the brain, the MS orchestrates oscillatory neuronal activities throughout the brain. These oscillations are critical in generating sensory and emotional salience, locomotion, maintaining mood, supporting innate anxiety, and governing learning and memory. Accumulating evidence points out that the physiological oscillations under septal influence are frequently disrupted or altered in pathological conditions. Therefore, the MS may be a potential target for treating neurological and psychiatric disorders with abnormal oscillations (oscillopathies) to restore healthy patterns or erase undesired ones. Recent studies have revealed that the patterned stimulation of the MS alleviates symptoms of epilepsy. We discuss here that stimulus timing is a critical determinant of treatment efficacy on multiple time scales. On-demand stimulation may dramatically reduce side effects by not interfering with normal physiological functions. A precise pattern-matched stimulation through adaptive timing governed by the ongoing oscillations is essential to effectively terminate pathological oscillations. The time-targeted strategy for the MS stimulation may provide an effective way of treating multiple disorders including Alzheimer’s disease, anxiety/fear, schizophrenia, and depression, as well as pain.

In the concept of the process signature, the relationship between a material load and the modification remaining in the workpiece is used to better understand and optimize manufacturing processes. The basic prerequisite for this is to be able to measure the loads occurring during the machining process in the form of mechanical deformations. Speckle photography is suitable for this in-process measurement task and is already used in a variety of ways for in-plane deformation measurements. The shortcoming of this fast and robust measurement technique based on image correlation techniques is that out-of-plane deformations in the direction of the measurement system cannot be detected and increases the measurement error of in-plane deformations. In this paper, we investigate a method that infers local out-of-plane motions of the workpiece surface from the decorrelation of speckle patterns and is thus able to reconstruct three-dimensional deformation fields. The implementation of the evaluation method enables a fast reconstruction of 3D deformation fields, so that the in-process capability remains given. First measurements in a deep rolling process show that dynamic deformations underneath the die can be captured and demonstrate the suitability of the speckle method for manufacturing process analysis.

High temperatures (HT) and high hydrostatic pressures (HHP) are characteristic of deep-sea hydrothermal vents and other deep crustal settings. These environments host vast and diverse microbial populations, yet only a small fraction of those populations have been successfully cultured. This is due, in part, to the difficulty of sampling while maintaining these in situ conditions and also replicating those high-temperature and high-pressure conditions in the laboratory. In an effort to facilitate more HT-HHP cultivation, we present two HT-HHP batch culture incubation systems for cultivating deep-sea vent and subsurface (hyper)thermophilic microorganisms. One HT-HHP system can be used for batch cultivation up to 110 MPa and 121°C, and requires sample decompression during subsampling. The second HT-HHP system can be used to culture microorganisms up to 100 MPa and 160°C with variable-volume, pressure-retaining vessels that negate whole-sample decompression during subsampling. Here, we describe how to build cost effective heating systems for these two types of high-pressure vessels, as well as the protocols for HT-HHP microbial batch cultivation in both systems. Additionally, we demonstrate HHP transfer between the variable-volume vessels, which has utility in sampling and enrichment without decompression, laboratory isolation experiments, as well as HHP filtration.

Hateful and abusive speech presents a major challenge for all online social media platforms. Recent advances in Natural Language Processing and Natural Language Understanding allow more accurate detection of hate speech in textual streams. This study presents a multimodal approach to hate speech detection by combining Computer Vision and Natural Language processing models for abusive context detection. Our study focuses on Twitter messages and, more specifically, on hateful, xenophobic and racist speech in Greek aimed at refugees and migrants. In our approach we combine transfer learning and fine-tuning of Bidirectional Encoder Representations from Transformers (BERT) and Residual Neural Networks (Resnet). Our contribution includes the development of a new dataset for hate speech classification, consisting of tweet ids, along with the code to obtain their visual appearance, as they would have been rendered in a web browser. We have also released a pre-trained Language Model trained on Greek tweets, which has been used in our experiments. We report a consistently high level of accuracy (accuracy score=0.970, f1-score=0.947 in our best model) in racist and xenophobic speech detection.

The evolution in imaging technologies and artificial intelligence algorithms, coupled with improvements in UAV technology, has enabled the use of unmanned aircraft in a wide range of applications. The feasibility of this kind of approach for cattle monitoring has been demonstrated by several studies, but practical use is still challenging due to the particular characteristics of this application, such as the need to track mobile targets and the extensive areas that need to be covered in most cases. The objective of this study was to investigate the feasibility of using a tilted angle to increase the area covered by each image. Deep Convolutional Neural Networks (Xception architecture) were used to generate the models for the experiments, which covered aspects like ideal input dimensions, effect of the distance between animals and sensor, effect of classification error on the overall detection process, and impact of physical obstacles on the accuracy of the model. Experimental results indicate that oblique images can be successfully used under certain conditions, but some practical limitations need to be addressed in order to make this approach appealing.

Carbon dioxide capture and storage (CCS) is highly expected to be mitigating the discharges of carbon dioxide in a surrounding environment. Solvents are an integral part of CCS. So far, several solvents have been explored the interest of meeting the requirements such as accessibility, non-harmfulness, biocompatibility, recyclability, and inexpensiveness. However, most solvents face failure in fulfilling the requirements due to many factors, so, this review paper gives a brief discussion about another category of solvent, an analogue of ionic liquids (ILs) named deep eutectic solvent (DES). Extensive research has been done on DES in recent years because of their various attractive advantages, i.e., non-poisonousness, biodegradability, cheap cost and easy preparation, that make them as a promising green solvent for many industrial procedure and application, for instance, polymer synthesis, biodiesel treatment, green chemistry, electrochemistry etc. Therefore, this manuscript mainly focusses on CO2 capturing through DES in oil and gas field. In addition, the preparation and chemical structures of this novel solvent (DES) is also discussed. Moreover, a detailed study based on experimental solubility of CO2 in DESs is also reported in this article.

Invasive species are significant threats to global agriculture and food security being the major causes of crop loss. An operative biosecurity policy requires full automation of detection and habitat identification of the potential pests and pathogens. Unmanned Aerial Vehicles (UAVs) mounted thermal imaging cameras can observe and detect pest animals and their habitats, and estimate their population size around the clock. However, their effectiveness becomes limited due to manual detection of cryptic species in hours of captured flight videos, failure in habitat disclosure and the requirement of expensive high-resolution cameras. Therefore, the cost and efficiency trade-off often restricts the use of these systems. In this paper, we present an invasive animal species detection system that uses cost-effectiveness of consumer-level cameras while harnessing the power of transfer learning and an optimised small object detection algorithm. Our proposed optimised object detection algorithm named Optimised YOLO (OYOLO) enhances YOLO (You Only Look Once) by improving its training and structure for remote detection of elusive targets. Our system, trained on the massive data collected from New South Wales and Western Australia, can detect invasive species (rabbits, Kangaroos and pigs) in real-time with a higher probability of detection (85–100 %), compared to the manual detection. This work will enhance the visual analysis of pest species while performing well on low, medium and high-resolution thermal imagery, and equally accessible to all stakeholders and end-users in Australia via a public cloud.

Face recognition is one of the most active research fields of computer vision and pattern recognition, with many practical and commercial applications including identification, access control, forensics, and human-computer interactions. However, identifying a face in a crowd raises serious questions about individual freedoms and poses ethical issues. Significant methods, algorithms, approaches, and databases have been proposed over recent years to study constrained and unconstrained face recognition. 2D approaches reached some degree of maturity and reported very high rates of recognition. This performance is achieved in controlled environments where the acquisition parameters are controlled, such as lighting, angle of view, and distance between the camera-subject. However, if the ambient conditions (e.g., lighting) or the facial appearance (e.g., pose or facial expression) change, this performance will degrade dramatically. 3D approaches were proposed as an alternative solution to the problems mentioned above. The advantage of 3D data lies in its invariance to pose and lighting conditions, which has enhanced recognition systems efficiency. 3D data, however, is somewhat sensitive to changes in facial expressions. This review presents the history of face recognition technology, the current state-of-the-art methodologies, and future directions. We specifically concentrate on the most recent databases, 2D and 3D face recognition methods. Besides, we pay particular attention to deep learning approach as it presents the actuality in this field. Open issues are examined and potential directions for research in facial recognition are proposed in order to provide the reader with a point of reference for topics that deserve consideration.

In the past years, various intelligent machine learning and deep learning algorithms have been developed and widely applied for gearbox fault detection and diagnosis. However, the real-time application of these intelligent algorithms has been limited, mainly due to the fact that the model developed using data from one machine or one operating condition has serious diagnosis performance degradation when applied to another machine or the same machine with a different operating condition. The reason for poor model generalization is the distribution discrepancy between the training and testing data. This paper proposes to address this issue using a deep learning based cross domain adaptation approach for gearbox fault diagnosis. Labelled data from training dataset and unlabeled data from testing dataset is used to achieve the cross-domain adaptation task. A deep convolutional neural network (CNN) is used as the main architecture. Maximum mean discrepancy is used as a measure to minimize the distribution distance between the labelled training data and unlabeled testing data. The study proposes to reduce the discrepancy between the two domains in multiple layers of the designed CNN to adapt the learned representations from the training data to be applied in the testing data. The proposed approach is evaluated using experimental data from a gearbox under significant speed variation and multiple health conditions. An appropriate benchmarking with both traditional machine learning methods and other domain adaptation methods demonstrates the superiority of the proposed method.

In this paper, our goal is to perform a virtual restoration of an ancient coin from its image. The present work is the first one to propose this problem, and it is motivated by two key promising applications. The first of these emerges from the recently recognised dependence of automatic image based coin type matching on the condition of the imaged coins; the algorithm introduced herein could be used as a pre-processing step, aimed at overcoming the aforementioned weakness. The second application concerns the utility both to professional and hobby numismatists of being able to visualise and study an ancient coin in a state closer to its original (minted) appearance. To address the conceptual problem at hand, we introduce a framework which comprises a deep learning based method using Generative Adversarial Networks, capable of learning the range of appearance variation of different semantic elements artistically depicted on coins, and a complementary algorithm used to collect, correctly label, and prepare for processing a large numbers of images (here 100,000) of ancient coins needed to facilitate the training of the aforementioned learning method. Empirical evaluation performed on a withheld subset of the data demonstrates extremely promising performance of the proposed methodology and shows that our algorithm correctly learns the spectra of appearance variation across different semantic elements, and despite the enormous variability present reconstructs the missing (damaged) detail while matching the surrounding semantic content and artistic style.

Deep brain stimulation (DBS) of the subthalamic nucleus (STN) has become an effective and widely used tool in the treatment of Parkinson’s disease (PD). STN-DBS has varied effects on speech. Clinical speech ratings suggest worsening following STN-DBS, but quantitative intelligibility, perceptual, and acoustic studies have produced mixed and inconsistent results. Improvements in phonation and declines in articulation have frequently been reported during different speech tasks under different stimulation conditions. Questions remain about preferred STN-DBS stimulation settings. Seven right-handed, native speakers of English with PD treated with bilateral STN-DBS were studied off medication at three stimulation conditions: stimulators off, 60 Hz (low frequency stimulation - LFS), and the typical clinical setting of 185 Hz (High frequency - HFS). Spontaneous speech was recorded in each condition and excerpts were prepared for transcription (intelligibility) and difficulty judgements. Separate excerpts were prepared for listeners to rate abnormalities in voice, articulation, fluency, and rate. Intelligibility for spontaneous speech was reduced at both HFS and LFS when compared to STN-DBS off. Speech produced at HFS was more intelligible than that produced at LFS, but HFS made the intelligibility task (transcription) subjectively more difficult. Both voice quality and articulation were judged to be more abnormal with STN-DBS on. STN-DBS reduced the intelligibility of spontaneous speech at both LFS and HFS but lowering the frequency did not improve intelligibility. Voice quality ratings with STN-DBS were correlated with the ratings made without stimulation. This was not true for articulation ratings. STN-DBS exacerbated an existing voice disorder and may have introduced new articulatory abnormalities.

Speech signals are degraded in real-life environments, product of background noise or other factors. The processing of such signals for voice recognition and voice analysis systems presents important challenges. One of the conditions that make adverse quality difficult to handle in those systems is reverberation, produced by sound wave reflections that travel from the source to the microphone in multiple directions.To enhance signals in such adverse conditions, several deep learning-based methods have been proposed and proven to be effective. Recently, recurrent neural networks, especially those with long and short-term memory (LSTM), have presented surprising results in tasks related to time-dependent processing of signals, such as speech. One of the most challenging aspects of LSTM networks is the high computational cost of the training procedure, which has limited extended experimentation in several cases. In this work, we present a proposal to evaluate the hybrid models of neural networks to learn different reverberation conditions without any previous information. The results show that some combination of LSTM and perceptron layers produce good results in comparison to those from pure LSTM networks, given a fixed number of layers. The evaluation has been made based on quality measurements of the signal's spectrum, training time of the networks and statistical validation of results. Results help to affirm the fact that hybrid networks represent an important solution for speech signal enhancement, with advantages in efficiency, but without a significan drop in quality.

With the increase of the Artificial Neural Network (ANN), machine learning has taken a forceful twist in recent times. One of the most spectacular kinds of ANN design is the Convolutional Neural Network (CNN). The Convolutional Neural Network (CNN) is a technology that mixes artificial neural networks and up to date deep learning strategies. In deep learning, Convolutional Neural Network is at the center of spectacular advances. This artificial neural network has been applied to several image recognition tasks for decades and attracted the eye of the researchers of the many countries in recent years as the CNN has shown promising performances in several computer vision and machine learning tasks. This paper describes the underlying architecture and various applications of Convolutional Neural Network.

We overview recent progress in growth aspects of group III-nitride heterostructures for deep ultraviolet (DUV) light-emitting diodes (LEDs), with particular emphasis on the growth approaches for attaining high-quality AlN and high Al-molar fraction AlGaN. The discussion commences with the introduction of the current status of group III-nitride DUV LEDs and the remaining challenges. This segues into discussion of LED designs enabling high device performance followed by the review of advances in the methods for the growth of bulk single crystal AlN intended as a native substrate together with a discussion of its UV transparency. It should be stated, however, that due to the high-cost of bulk AlN substrates at the time of this writing, the growth of DUV LEDs on foreign substrates such as sapphire still dominates the field. On the deposition front, the heteroepitaxial growth approaches incorporate high-temperature metal organic chemical vapor deposition (MOCVD) and pulsed-flow growth, a variant of MOCVD, with the overarching goal of enhancing adatom surface mobility, and thus epitaxial lateral overgrowth which culminates in minimization the effect of lattice- and thermal-mismatches. This is followed by addressing the benefits of pseudomorphic growth of strained high Al-molar fraction AlGaN on AlN. Finally, methods utilized to enhance both p- and n-type conductivity of high Al-molar fraction AlGaN are reviewed.

Manipulating cloth-like deformable objects (CDOs) is a long-standing problem in the robotics community. CDOs are flexible (non-rigid) objects that do not show a detectable level of com- pression strength while two points on the article are pushed towards each otherand include objects such as ropes (1D), fabrics (2D) and bags (3D). In general, CDOs’ many degrees of freedom (DoF) introduce severe self-occlusion and complex state-action dynamics as significant obstacles for perception and manipulation systems. These challenges exacerbate existing issues of modern robotic control methods such as imitation learning (IL) and reinforcement learning (RL). This review focuses on the application details of data-driven control methods on four major task families in this domain: cloth-shaping, rope manipulation, dressing and bag manip- ulation. Furthermore, we identify specific inductive biases in these four domains that present challenges for more general IL and RL algorithms, and summarise the future direction for the development of the field.

The paper presents a study on optimizing leather defect detection using the YOLOv11 model. Traditional leather processing has faced challenges in quality control, where human inspection accuracy ranges between 70% to 85%, affecting leather utilization rates and resulting in significant material waste. This research introduces an automated solution to improve defect classification and detection accuracy. The study focuses on defects commonly found in leather, such as insect larvae damage and removal cuts, using a specialized light chamber to control environmental variables. By analyzing both grain and flesh sides of the leather, the researchers demonstrated a significant increase in defect detection accuracy, with the flesh side achieving a higher classification rate. YOLOv11's dual-side analysis allowed for clearer identification of subtle defects, leading to a more efficient classification process. The results suggest that integrating advanced AI models like YOLOv11 with controlled digitization environments could drastically reduce human error and improve leather utilization, providing a scalable solution for the leather industry.

This paper presents a novel machine learning framework that applies evolutionary ensembles of artificial agents to mineralogical analysis and classification. The approach is based on hybridiza-tion techniques that combine diverse machine learning algorithms, creating large and effective communities of agents. These progressively mute and improve through crossover, hybridization, and selection, addressing the challenges of mineral recognition and classification from thin section images. By combining multiple machine learning techniques, the ensemble of agents autono-mously improves by evolving to adapt and enhance its ability to identify mineral species and classify different types of alterations. We detail the method, provide examples using synthetic and real data, and explore the potential to improve mineralogical analysis workflows through this dynamic, self-improving system.

The evaluation of wrinkles in deep-drawn fiber-based materials is crucial for the assessment of product quality and the optimization of manufacturing processes. Wrinkles, a common phenomenon in deep-drawing paper materials with high drawing depths, significantly impair the appearance and mechanical properties of the final product. The objective of this study is to identify the key process parameters affecting wrinkling and to deepen the understanding of their roles and interactions using wrinkle data for deep-drawn paper products. Image analysis techniques are employed, supported by a specially constructed darkroom platform to ensure uniform light intensity for capturing photographs. An automated program is developed for the detection and evaluation of wrinkle characteristics and distribution, which allows the free choice of the region to be detected and the representation of the wrinkle geometry not limited by the number. To enhance the precision of this program, the ellipticity is initially rectified for products without flanges, specifically cup-shaped deep-drawn products. The ellipticity is caused by the pronounced springback effect of the paperboard. The approach is employed to investigate the impact of material properties, blank holder force, drawing depth, drawing clearance, and punch speed on wrinkling formation after the deep drawing process. The findings reveal that the blank holder force and drawing clearance are critical factors in wrinkle formation, with higher forces generally leading to increased wrinkle numbers.

Robot grasp modeling and implementation is one of essential abilities for a robot with embodied artificial intelligence. However, most existing deep learning-based grasp methods suffer a large number of parameters and heavy computation overheads. To address this issue, by fully exploiting the complementary capabilities of both CNNs and Transformers, we propose a smart grasp deep network with one-way fusion strategy via context path and spatial path (SGNet), which enjoys a lightweight structure, fast inference speed and easy deployment on devices with limited computation resources. Specifically, the context path employs lightweight depthwise separable convolution to achieve fast down-sampling while a novel DSFormer module mainly by integrating Transformer is to extract global and context-rich features. The spatial path efficiently fuses feature information from the context path in one-way manner and generate high-resolution feature maps via point-by-point convolution operations. Experimental results show the proposed model with only 1M parameters has a significantly overall performance, achieving 99.4% accuracy on Cornell dataset and 93.4% accuracy on Jacquard dataset, as well as within 12.5ms inference time.

Heart rate and blood pressure, along with body temperature and respiratory rate, are two of the four parameters used to assess the state of bodily functions, as both measures are indicators of heart activity. Therefore, these measurements become a basic procedure in patient diagnosis, necessitating techniques that can obtain them accurately and reliably, while being as non-intrusive as possible for the patient. Using video data, these two vital signs can be determined at any time and place. To achieve this, artificial intelligence techniques such as machine learning and deep learning 6 can be applied to overcome the main limitations of conventional remote photoplethysmography (rPPG). This document explains the technical and methodological aspects involved, including the acquisition and preprocessing of video data, the extraction of the rPPG signal, and its subsequent processing, as well as the artificial intelligence model for heart rate estimation and the mathematical modeling for blood pressure estimation. We compared the results for heart rate and blood pressure extracted from videos recorded by a basic smartphone front-camera to a blood pressure monitor and achieved results with reasonable accuracy and correlation. These results are part of the AIVA project, grant agreement 2021/C005/00145071. This project has been funded by Red.es through the Spanish 2021 call for grants aimed at research and development projects in artificial intelligence and other digital technologies.

Detecting people and technical objects in various situations, such as natural disasters and warfare, is critical to search and rescue operations and the safety of the civilian. A fast and accurate detection for people and equipment can significantly increase the effectiveness of search and rescue missions and provide timely assistance to people. Computer vision and deep learning technologies play a key role in detecting the required objects due to their ability to analyse big volumes of visual data in real time. The performance of the neural networks such as YOLOv4-v8, Faster R-CNN, SSD, and EfficientDet has been analysed using COCO2017, SARD, SeaDronesSee, and VisDrone2019 datasets. Main criteria for comparison were mAP, Precision, Recall, F1-Score, and the ability of the neural network to work in real time. The most important metrics for evaluating the efficiency and performance of models for a given task are accuracy (mAP), F1-Score, and processing speed (FPS). These metrics allow us to evaluate both the accuracy of object recognition and the ability to use the models in real-world environments where high processing speed is important. Although different neural networks perform better on certain types of metrics, YOLO outperforms them on all metrics, showing the best results of mAP-0.88, F1 - 0.88, and FPS - 48, so the focus was on these models.

The growing use of drones in precision agriculture highlights the need for enhanced operational efficiency. Despite the ability of computer vision based on deep learning has made remarkable progress in the past ten years, when it comes to segmentation task on UAVs, there is always a conflict between the demand of high precision and low inference latency. Due to such a dilemma, we propose the IRB-YOLO, an efficiency model based on Inverted Residual Block, devoting to provide constructive strategies in real-time detection tasks of UAV camera. The working details of this paper are as follows: (1) This paper innovates with a IR-Block(Inverted Residual Block), integrated into a refined YOLOv8-seg structure to create IRB-YOLO. This model specializes in pixel-level classification of UAV-acquired RGB images, facilitating the creation of exact maps to guide agricultural strategies. (2)When it comes to the experiments on a Vatica dataset with any other light-weight segmentation model, IRB-YOLO achieve at least a 3.3% increase in mAP. Further validation using a diverse species dataset confirms its robust generalization. (3)Without overloading the complex attention mechanism and deeper and deeper network, a stem that incorporates efficient feature extraction components, inverted residual block, can still possess outstanding modeling capabilities. IRB-YOLO builds a bridge between academic research and edge deployment of drones, making it applicable in real-world scenarios.

Drought and water scarcity are significant factors limiting the development of China's maize industry. Deep sowing, as a method to mitigate drought and preserve soil moisture and seedlings, can effectively mitigate the adverse effects of drought stress on their growth. However, deep sowing can influence the emergence of maize seedlings and has a certain impact on their growth and development. To investigate the physiological and transcriptomic changes in maize inbred lines under deep sowing stress, and the alleviation effects of exogenous hormones, this study investigated the response of mesocotyl in deep-seeding tolerant maize inbred line Qi319 and deep-seeding sensitive maize inbred line Zi330 to deep-seeding stress under exogenous MeJA treatment. The physiological and transcriptomic mechanisms underlying the alleviation of deep-seeding stress by exogenous MeJA were elucidated. Results showed that compared to normal sowing depth, mesocotyl length (MES), mesocotyl and coleoptile length (MESCOL) significantly increased under deep-seeding stress for both inbred lines, while seedling length (SDL) was significantly inhibited. However, after adding 1.5 μmol·L^-1 MeJA, MES, MESCOL, and SDL significantly increased compared to deep-seeding stress alone. Exogenous MeJA promoted cell elongation, enhanced mesocotyl length, and improved seedling emergence ability. RNA-Seq sequencing revealed differentially expressed genes (DEGs). Under deep-seeding stress compared to normal sowing depth conditions, 1248 DEGs were identified in Qi319 and 1288 DEGs were identified in Zi330. After adding exogenous MeJA under deep-seeding stress compared to normal sowing depth with exogenous MeJA treatment conditions, 538 DEGs were identified in Qi319 and 2284 DEGs were identified in Zi 330. GO enrichment analysis showed that differentially expressed genes were mainly concentrated in biological processes, cell components and molecular functions. KEGG enrichment analysis showed that differentially expressed genes were mainly concentrated in transport and catabolism, signal transduction, amino acid metabolism and lipid metabolism processes. Compared with the deep-seeded treatment, 12 differentially expressed genes were co-expressed in the two inbred lines under exogenous MeJA treatment, and 10 key genes, including ornithine decarboxylase, terpene synthase 7 and ethylene reaction transcription factor 11, were screened out. Our findings elucidate the molecular mechanism underlying maize's tolerance response to deep seeding and offer valuable insights into candidate gene resources for improving deep seeding tolerance in maize.

Crawling robots are the focus of intelligent inspection research, and the main feature of this type of robot is the flexibility of in-plane attitude adjustment. The crawling robot HIT_Spibot is a new type of steam generator heat transfer tube inspection robot with a unique mobility capability different from traditional quadrupedal robots. This paper introduces a hierachical motion planning approach for HIT_Spibot, aiming to achieve efficient and agile maneuverability. The proposed method integrates three distinct planners to handle complex motion tasks: a nonlinear optimization-based base motion planner, a TOPSIS-based base orientation planner, and a Mask-D3QN (MD3QN) algorithm-based gait motion planner. Initially, the robot's base and foot workspace are delineated through envelope analysis, followed by trajectory computation using Larangian methods. Subsequently, the TOPSIS algorithm is employed to establish an evaluation framework conducive to foundational turning planning. Finally, the MD3QN algorithm trains foot points to facilitate robot movement along predefined paths. Experimental results demonstrate the method's adaptability across diverse tube structures, showcasing robust performance even in environments with random obstacles. Compared to the D3QN algorithm, MD3QN achieves a 100% success rate, enhances average overall scores by 6.27%, reduces average stride lengths by 39.04%, and attains a stability rate of 58.02%. These results not only validate the effectiveness and practicality of the method but also showcase the significant potential of HIT_Spibot in the field of industrial inspection.

Increasing concerns on animal welfare in commercial pig industry include aggression between pigs as it affects their health and growth. Early detection of aggressive behaviors is essential for optimizing their living environment. A major challenge for detection is that these behaviors are observed occasionally in normal conditions. Under this circumstance, a limited amount of aggressive behavior data will lead to class imbalance issue, making it difficult to develop an effective classification model for the detection of aggressive behaviors. To overcome this problem, we approach the aggressive behavior detection problem as a case of anomaly detection rather than classification and propose a model based on anomaly detection, which can better handle unbalanced class distribution and effectively detect the infrequent, aggressive episodes of pigs. The model consists of a convolutional neural network (CNN) and a variational long short-term memory (LSTM) autoencoder. Additionally, we adopted a training method similar to weakly supervised anomaly detection and included a few aggressive behavior data in the training set for prior learning. To effectively utilize the aggressive behavior data, we created Reconstruction Loss Inversion, a novel objective function, to train the autoencoder-based model to increase the reconstruction error of aggressive behaviors by inverting the loss function. Our anomaly detection approach significantly outperforms traditional classification-based methods, effectively identifying aggressive behaviors in a natural farming environment. This method offers a robust solution for detecting aggressive animal behaviors and contributes to improving their welfare.

Large-scale deep learning models have produced significant results in underwater visual image target detection. However, deploying these models on underwater embedded devices poses a series of issues and challenges. Firstly, underwater devices have limited communication capabilities and carry relatively short supply of energy, resulting in constrained computing power for edge devices. Secondly, the large scale of deep neural networks creates a conflict in power demands with unmanned underwater equipment. Thirdly, insufficient underwater light and turbid water quality will lead to degradation problems in the images acquired by the device. In this paper, we build an offline system to complete the task of underwater scene target recognition. This paper introduces several innovations: 1) The new YOLO-TN model is developed by compressing the target recognition model based on YOLO-V5. The model use a model compression algorithm replace the YOLO-V5’s backbone network with a lightweight network, then perform parameter pruning on the model detection head, while reduce the size of the input image. The model uses a Teacher-guided Neural Architecture Search method based on YOLO-V5, named as YOLO-TN. 2)A method for constructing and processing real underwater environment data sets is designed. The datasets are all captured using underwater unmanned vehicles in real underwater environments, addressing issues present in existing underwater datasets, such as unbalanced target volume and single-image environment. Experimental results demonstrate that the YOLO-TN model significantly reduces the model size while maintaining the original recognition accuracy. Furthermore, it achieves 28.6 FPS on embedded devices, which is 12 times that of the original model.

As computer networks develop in size and complexity, the need for equipment that can characterize, inspect, monitor, and summarize the features of a transmission in real time grows. Many transmission and packet inspection theories are required to examine network traffic. Unfortunately, loss the track or traceability of one traffic scenarios against others over real-time network. Deep packet inspection (DPI) approaches can be used to alleviate the problem but still face flexibility constraints in the sense that it will be difficult to keep track of prior occurrences. There have been significant theoretical improvements in packet inspection, but the problem of event tracking flexibility remains unsolved. A novel approach for “integrating blockchain with DPI” in summarizing computer network traffic flow events has emerged as a result of this research: This technique captured traffic flows in the network, DPI report, and the entire transmission session header and load them into a smart-contract based- Blockchain. Two experimental scenarios are used to test the validity of the system. The first scenarios involve analysis of DPI transmission without integrating blockchain, follow by the captured of transmission session involving DPI with blockchain. The finding indicates that DPI transmission with blockchain is more secured against "Tampering Incidents", "Unauthorized Access Attempts", “Auditability” and many other security concern. Whereas, the performance overhead introduced by blockchain-based DPI indicate an increase in "DPI Processing Time from 68ms to 150ms, a "CPU Usage increment by 20%, Network Latency increment of 15ms from 20ms to 35ms, and almost all other computational resources increase. Therefore, the finding can be summarized that the performance overhead introduced by integrating blockchain with DPI increases, but it provides a secure DPI reports for decision making.

Deep brain stimulation of the subthalamic nucleus (STN-DBS) is a recognized treatment in Parkinson’s disease (PD), but knowledge is still limited regarding its possible impact on patients with behavioral addiction (BA) who undergo surgery. To assess this effect of STN-DBS, we prospectively studied the follow-up from 3-months pre-surgery to 12-months post-surgery of 12 PD patients with pre-existing BA (9 with pathological gambling disorder, 2 with hypersexual disorder, and 1 with compulsive shopping disorder) in a multicenter study. Patients were assessed for BA, for mental status with anxiety, depression, mania, emotional process and for PD, using the Unified Parkinson’s Disease Rating Scale motor examination (UPDRS3) and cumulative daily L-dopa equivalent daily dose (LEDD). Mean 6 and 12-month post-surgery scores were compared to baseline. After one year of STN-DBS, our results indicated that the treatment remained generally effective and stable. There was no effect on mania state or negative impact on motor signs. A significant improvement of the anxiety score was also observed. The STN is an interesting target for new therapeutic stimulation in BA and works to investigate the precise target in this neurobiological architecture should be performed in the future.

The rapidly evolving domain of multimodal sentiment analysis is crucial for unraveling the intricate layers of emotional expression in social media content, customer service interactions, and personal vlogs. This research introduces a cutting-edge Advanced Cross-Modal Gating (ACMG) framework that significantly enhances the precision of sentiment analysis by refining the interplay among textual, auditory, and visual modalities. Our approach addresses three foundational aspects of sentiment analysis: (1) Advanced learning of cross-modal interactions, which focuses on extracting and synthesizing sentiment from varied modal inputs, thus providing a holistic view of expressed emotions; (2) Mastery over the temporal dynamics of multimodal data, enabling the model to maintain context and sentiment continuity over extended interactions; and (3) Deployment of a novel fusion strategy that not only integrates unimodal and cross-modal cues but also dynamically adjusts the influence of each modality based on its contextual relevance to the sentiment being expressed. The exploration of these dimensions reveals that the nuanced modeling of cross-modal interactions is crucial for enhancing model responsiveness and accuracy. By applying the ACMG model to two highly regarded datasets—CMU Multimodal Opinion Level Sentiment Intensity (CMU-MOSI) and CMU Multimodal Opinion Sentiment and Emotion Intensity (CMU-MOSEI)—we achieve groundbreaking accuracies of 83.9\% and 81.1\%, respectively. These results represent significant improvements of 1.6\% and 1.34\% over the current state-of-the-art, showcasing the superior performance and potential of our approach in navigating the complexities of multimodal sentiment analysis.

Colorectal cancer is third most common malignancy. For metastasis of rectal cancer, metastasis to skeletal muscle tissue is rare. In this case report, we describe a 51-year-old male patient with rectal cancer with neoadjuvant concurrent radiotherapy and chemotherapy followed by transanal total mesorectal excision with ileostomy. However, he had deep venous thrombosis of left leg after operation and skeletal muscle metastasis was diagnosed via tissue proof. Therefore, less possibility of metastasis increased included brain, duodenum, bone, and nervous system. The prognosis of metastasis to skeletal muscle is poor in contrast to liver or lung metastasis.

The aim of the present study was to predict the response of non-small cell lung cancer (NSCLC) patients to immune checkpoint inhibitors (ICIs) by leveraging computed tomography (CT) images using deep learning techniques. Retrospectively, 624 sequential CT images were gathered from 156 patients at Jiangsu Province Hospital, along with their clinical data. The dataset was subsequently partitioned into three groups: training (n=547), validation (n=64), and test (n=64). Moreover, an external validation cohort included 37 CT images from patients at Nanjing Pukou Peoples' Hospital, accompanied by comprehensive clinical data. An advanced Video Vision Transformer (ViViT) model incorporating global self-attention was utilized to analyse patients treated with ICIs and predict their response. The ViViT model's efficacy was evaluated using a confusion matrix and a receiver operating characteristic curve (ROC). Notably, the ViViT model demonstrated predictive prowess for ICIs response, yielding respective areas under the receiver operating characteristic curve (AUC) of 0.74 (95% CI: 0.69-0.78), 0.74 (95% CI: 0.61-0.86), 0.76 (95% CI: 0.62-0.88), and 0.69 (95% CI: 0.5-0.87) in the training, validation, test, and external validation cohorts. The present study illustrates how a deep learning model can provide a non-invasive means to predict clinical outcomes in NSCLC patients undergoing ICIs, potentially transforming personalized treatment approaches for individuals with NSCLC.

This study evaluates the effectiveness of Darknet53 and VGG16 pretrained models to develop a hybrid ensemble model named HybridPlantNet23 deep learning model for effective plant disease classification, using the publicly available dataset ‘A Dataset of Leaf Images’ from Mendeley webpage. The preprocessing of the dataset involved resizing the images to the respective requirements of the pretrained models to develop a combined datastore which served as the input. All experiments were conducted using MATLAB R2023b deep network designer interface. Model performances were evaluated based on the conventional accuracy, precision, recall, and F1-score. VGG16 achieved the highest accuracy (97.25%), precision (96.30%), recall (96.96%), and F1-score (96.55%), outperforming both Darknet53 and the ensemble HybridPlantNet23 model. Despite the advanced architecture of the ensemble model, its performance (accuracy: 95.37%, precision: 93.56%, recall: 94.20%, F1-score: 93.74%) was not significantly superior, thus highlighting scientific reasons behind this development. Recommendations for future works were provided accordingly.

Objectives: To assess the concordance of the preoperative application of the #Enzian classification (#ENZIANi) with the postoperative result (#ENZIANs) using surgical findings as the reference standard. Methods: This retrospective study included 282 consecutive patients with deep endometriosis undergoing surgical treatment. Preoperative assessment with transvaginal sonography and magnetic resonance imaging was compared with postoperative assessment. Concordance and diagnostic test evaluation were calculated. Results: The highest concordance was observed in the F (abdominal wall endometriosis) with k Cohen of 0.837, following the values for pelvic locations, with 0.795 for T left, 0.791 for T right, 0.776 for F (adenomyosis), 0.766 for C (rectum), and 0.75 and 0.72 for O right k and O left, respectively. The highest sensitivity was demonstrated for the P compartment *(98%), T compartment (both sides 97%), and A, B, C (94%- 96%), corresponding with deep endometriosis. Conclusions: Preoperative assessment using TVS/TAS+MRI with the ENZIANi score correlates well with the ENZIANs postoperative score and demonstrates good concordance in the detection and localization of deep endometriosis, thereby minimizing false negative results and ensuring accurate preoperative staging. The ENZIAN classification is well-suited to surgeons' needs and benefits from continuous development. Future improvements, such as adding the expanded C module, may be considered in the next edition.

Prochlorococcus, a cyanobacteria genus of the smallest and most abundant oceanic phototrophs, encompasses ecotype strains adapted to high-light (HL) and low-light (LL) niches. To elucidate the adaptive evolution of this genus, we analyzed 40 Prochlorococcus marinus ORFeomes, including two cornerstone strains, MED4 and NATL1A. Employing deep learning with robust statistical methods, we detected new protein family distributions in the strains and identified key genes differentiating HL and LL strains. HL strains harbor genes (ABC-2 transporters) related to stress resistance, such as DNA repair and RNA processing, while LL strains exhibit unique chlorophyll adaptations (ion transport proteins, HEAT repeats). We report variable, depth-dependent endogenous viral elements in the 40 strains. We constructed the ORFeomes of MED4 and NATL1A covering 99 % of the annotated protein-coding sequences of the two species, totaling 3976 cloned, sequence-verified ORFs. These comparative genomics analyses, paired with MED4 and NATL1A ORFeomes, will facilitate future genotype-to-phenotype mappings and systems biology exploration of Prochlorococcus ecology.

Deep eutectic solvents have emerged as a greener alternative and have attracted so much interest in the last two decades. This paper present alternative synthesis method to classical heating-stirring. The ultrasound method is one of the most promising syntheses DESs in terms of yield and energy. So, ultrasound synthesis method was studied to obtain hydrophobic (Aliquat 336: L-Menthol 3:7; Li-docaine: decanoic acid 1:2) and hydrophilic DESs based on Choline chloride and urea, ethylene glycol and oxalic acid. The physical characterization of DESs by comparison of FTIR spectra no show difference between DESs obtained by heating-stirring and ultrasound synthesis methods. The RMN spectra of the DESs obtained by heating-stirring revealed the DESs formation. Density and viscosity properties of DESs were evaluated. The density values are similar for both synthesis methods. However, lower viscosity values are obtained to the hydrophilic DESs prepared by ul-trasound methods.

A surprising insight from the advent of genomic sequencing was that many genes are deeply conserved during evolution. With a particular focus on genes that interact with light in animals, I explore the metaphor of genetic toolkits, which can be operationalized as lists of genes involved in a trait of interest. A fascinating observation is that genes of a toolkit are often used again and again during convergent evolution, sometimes across vast phylogenetic distances. Such a pattern in the evolution of toolkits requires three different stages: origin, maintenance, and redeployment of the genes. The functional origins of toolkit genes might often be rooted in interactions with external environments. The origins of light interacting genes in particular may be tied to ancient responses to photo-oxidative stress, inspiring questions about the extent to which the evolution of other toolkits were also impacted by stress. Maintenance of genetic toolkits over long evolutionary timescales requires gene multifunctionality to prevent gene loss when a trait of interest is absent. Finally, the deployment of toolkit genes in convergently evolved traits like eyes sometimes involves the repeated use of similar, ancient genes but other times involves different genes specific to each convergent origin. How often a particular gene family is used time and again for the same function may depend on how many possible biological solutions are available. When few solutions exist and are maintained, evolution is constrained to use the same genes over and over. However, when many different solutions are possible, the innovative possibilities of evolution are often on display. Therefore, a focus on genetic toolkits highlights the combination of legacy plus innovation that drives the evolution of biological diversity.

Underwater visual detection technology plays a pivotal role in fields such as marine exploration. With the increasing demand for underwater monitoring, the quest for efficient and reliable methods for underwater target recognition has become particularly significant. To address this requirement, this study developed an innovative underwater object detection architecture based on YOLOv8, named YOLOv8-MU, aimed at significantly enhancing detection accuracy.By integrating the LarK module proposed in UniRepLKNet to optimize the backbone network, YOLOv8-MU aims to achieve a larger receptive field without increasing the model’s depth. Further, this research introduces C2fSTR, an innovative method that combines Swin Transformer with the C2f module. Additionally, we have incorporated the SPPFCSPC_EMA module, which combines Cross-Stage Partial Fast Spatial Pyramid Pooling (SPPFCSPC) with attention mechanisms, significantly improving the detection accuracy and robustness of various biological targets. Moreover, by introducing a fusion block based on DAMO-YOLO into the neck of the model, we further enhanced the model’s capability in multi-scale feature extraction. Finally, the adoption of the MPDIoU loss function, designed around vertex distance, effectively tackles the challenges of localization accuracy and boundary clarity in underwater organism detection. Experimental results on the URPC2019 dataset demonstrate that the YOLOv8-MU model achieved an mAP@0.5 of 78.4%, marking improvements of 5.6%, 1.1%, and 4.0% over YOLOv5s, YOLOv7, and YOLOv8n respectively, indicating the leading performance (SOTA) of this method. On the other hand, further evaluation on the URPC2020 dataset confirmed the generalization capability of the YOLOv8-MU architecture, with its mAP@0.5 reaching 80.4%, surpassing various models including YOLOv5x and YOLOv8n, showcasing the wide applicability of our proposed improved model architecture.

The recently explored synergistic combination of graphene-based materials and deep eutectic solvents (DESs) is opening novel and effective avenues for developing sensing devices with optimized features. In more detail, remarkable potential in terms of simplicity, sustainability, and cost-effectiveness have been demonstrated for sensors, resulting in the creation of hybrid devices with enhanced signal-to-noise ratio, linearity, and selectivity. Therefore, this review aims to provide a comprehensive overview of the currently available scientific literature discussing investigations and applications of sensors that integrate graphene-based materials and deep eutectic solvents, with an outlook for the most promising developments of this approach.

Despite UAV multi-target detection exhibits considerable developmental potential worldwide, it suffers distinct challenges compared with traditional tasks in this field. These challenges include insufficient feature extraction capabilities for small targets, limited capabilities for multi-dimensional feature fusion, as well as constraints on hardware computation parameters. Especially in mission scenarios such as disaster detection, these challenges will be further amplified. Consequently, this paper introduces SSE-YOLO, an innovative YOLO framework algorithm specially designed to address these challenges. To enhance the model's feature extraction capability, we employ the SPDConv module to replace the original Conv in the backbone section, utilizing depth-separable convolution instead of traditional convolution pooling. Concurrently, we eliminate the SPPF module at the bottom and address a new Separate Kernel Attention Pyramid Pooling (SKAPP) module, substantially enhancing the feature fusion capability at the model's core. Moreover, to address the challenge of multi-dimensional feature fusion, we replace the Concat module of the neck and head with E-BiFPN, transmitting feature information from the backbone to the lower network through four CBS blocks, which effectively resolves the issue of lost contextual information in the model. Meanwhile, SSE-YOLO undergoes ablation experiments on the VisDrone2019 dataset to evaluate its effectiveness against alternative methods, and experimental results illustrate the model's exceptional precision in detecting UAV targets. In comparison to models with comparable experimental accuracy, SSE-YOLO requires remarkably fewer parameters. On the VisDrone2019-DET-test-dev dataset, SSE-YOLO enhances mAP by 17.3%, with a 42.5% reduction in the parameter amounts. Therefore, the proposed method effectively tackles the challenge of reconciling low parameters and high accuracy, presenting a novel pathway for deep learning-based UAV multi-target detection.

To ensure the security of highly sensitive remote sensing images (RSI) during their transmission, it is essential to implement effective content security protection methods. Generally, secure distribution schemes for remote sensing images often employ cryptographic techniques. However, sending encrypted data exposes communication behavior, which poses significant security risks to the distribution of remote sensing images. Therefore, this paper introduces deep information hiding to achieve secure distribution of remote sensing images, which can serve as an effective alternative in certain specific scenarios. Specifically, the Deep Information Hiding for RSI Distribution (hereinafter referred to as DIH4RSID) based on encoder-decoder network architecture with Parallel Attention Mechanisms (PAM) by adversarial training was proposed. Our model is constructed with four main components: a preprocessing network (PN), an embedding network (EN), a revealing network (RN) and a discriminating network (DN). The PN module is primarily based on Inception to capture more details of RSI and targets of different scales. The PAM module obtains features in two spatial directions to realize feature enhancement and context information integration. The experimental results indicate that our proposed algorithm achieves relatively higher visual quality and secure level compared to related methods. Additionally, after extracting the concealed content from hidden images, the average classification accuracy is unaffected.

To support pathologists in breast tumor diagnosis, deep learning plays a crucial role in the development of histological whole slide image (WSI) classification methods. However, automatic classification is challenging due to the high-resolution data and the scarcity of representative training data. To tackle these limitations, we propose a deep learning-based breast tumor gigapixel histological image multi-classifier integrated with a high-resolution data augmentation model to process the entire slide by exploring its local and global information and generating its different synthetic versions. The key idea is to perform classification and augmentation in feature latent space, reducing the computational cost while preserving the class label of the input. We adopt a deep learning-based multi-classification method and evaluate the contribution given by a Conditional Generative Adversarial Network-based data augmentation model on the classifier’s performance for three tumor classes in the BRIGHT Challenge dataset. The proposed method has allowed us to achieve an average F1 equal to 69.5, considering only the WSI dataset of the Challenge. The results are comparable to those obtained by the Challenge winning method (71.6), also trained on the annotated tumor region dataset of the Challenge.

This paper introduces an innovative approach to lipreading, leveraging a web application designed to generate subtitles for videos where the speaker's mouth is visible and a comprehensive literature review that precedes the discussion, encompassing a thorough examination of various lipreading methods employed over the past decade. Our method employs a powerful deep learning model, featuring a 3D-convolution network and bidirectional LSTM, enabling accurate sentence-level predictions based solely on visual lip movements. With an impressive accuracy of 97%, our model is trained using pre-segmented lips regions, transformed into animated GIFs for effective pre-training. This work stands as a significant contribution to the evolving landscape of lipreading research, offering a practical and accurate solution for real-world applications.

With the development of remote sensing satellite technology for Earth observation, remote sensing stereo images have been used for three-dimensional reconstruction in various fields, such as urban planning and construction. However, remote sensing images often include noise, occluded regions, weakly textured areas and repeated textures, which can lead to reduced accuracy in stereo matching and affect the quality of the 3D reconstruction results. To reduce the impact of complex scenes in remote sensing images on stereo matching and to ensure both speed and accuracy, we propose a new end-to-end stereo matching network based on convolutional neural networks (CNNs). The proposed stereo matching network can learn features at different scales from the original images and construct cost volumes with varying scales to obtain richer scale information. Additionally, when constructing the cost volume, we introduce negative disparity to adapt to the common occurrence of both negative and nonnegative disparities in remote sensing stereo image pairs. For cost aggregation, we employ a 3D convolution-based encoder-decoder structure that allows the network to adaptively aggregate information. Before feature aggregation, we also introduce an attention module to retain more valuable feature information, enhance feature representation, and obtain a higher-quality disparity map. By training on the publicly available US3D dataset, we obtain the accuracy that 1.115 pixel in end-point error (EPE) and 5.32% in the error pixel ratio (D1) on the test dataset, and the inference speed is 92 ms. Comparing our model with existing state-of-the-art models, we achieve higher accuracy, and the network is beneficial for the three-dimensional reconstruction of remote sensing images.

Articular cartilage lacks intrinsic regenerative capabilities, and the current treatments fail to regenerate damaged tissue and lead only to temporary pain relief. These limitations have prompted the development of tissue engineering approaches, including 3D culture systems. Thanks to their regenerative properties and capacity to recapitulate embryonic processes, spheroids obtained from mesenchymal stromal cells are increasingly studied as building blocks to obtain functional tissues. The aim of this study was to investigate the capacity of adipose stromal cells to assemble in spheroids and differentiate toward chondrogenic lineage from the perspective of cartilage repair. Spheroids were generated by two different methods (3D chips vs Ultra-Low Attachment plates), differentiated toward chondrogenic lineage, and their properties were investigated by: molecular biology analyses, biophysical measurement of mass density, weight, and size of spheroids, and confocal imaging. Overall, spheroids showed the ability to differentiate by expressing specific cartilaginous markers that correlate with their mass density, defining a critical point at which they start to mature. Considering the spheroid generation method, this pilot study suggested that spheroids obtained with chips are a promising tool for the generation of cartilage organoids that could be used for preclinical/clinical approaches, including the realization of personalized therapy.

Real-time tracking is one of the most challenging problems in computer vision. Most Transformer-based trackers usually require expensive computational and storage power, which leads to the inability of these robust trackers to achieve real-time performance in resource-constrained devices. In this work, we propose a lightweight tracker, AnteaTrack. To localize the target more precisely, we develop a scaling-invariant max-filtering operator employing a sliding window combined with local max-pooling, which filters out the suspected target from the feature and performs an augmented representation while suppressing the background. For a more tight bounding box, we employ Pixel-Shuffle to increase the resolution of the feature map and get a more fine-grained representation of the target. In addition, AnteaTrack can run in real-time at 47 frames per second(FPS) on the CPU. We tested AnteaTrack on 5 datasets, and a large number of experiments have shown that AnteaTrack provides the most efficient solution compared to the same type of CPU real-time trackers. The code will be available at https://github.com/cnchange/AnteaTrack.

Software Defined Networking (SDN) is positioning the standard for the management of networks due to its scalability and flexibility to program the network. The SDN provides many advantages but it also involves some specific security problems take down the controller using cyber attack and in result the whole network will shut down which makes it a single point of failure. In this paper, the DDoS attacks in SDN were detected using AI-enabled, machine and deep learning, models with some specific features for data-set under normal and DDoS traffic. In our approach, initial data-set is collected from 84 features on kaggle and then the 20 top most features are selected using permutation importance algorithm. The data-set were learned and tested with AI-enabled 5 models. Our experimental results showed that the use of machine learning based random forest model has achieved the highest accuracy rate of 99.97%, in DDoS attack detection in SDN. Our contributions through this study are, first, we found highest 20 attacks that absolutely contributed to DDoS attacks. Secondly, it can reduce the time and cost of comparing various learning models and performance required for determining a learning model suitable for DDoS detection. Finally, various experimental methods for evaluating the performance of the learning model are presented so that related researchers can utilize them.

Due to the complex underground environment, pumping machines are prone to produce numerous failures. The indicator diagrams of faults are similar in a certain degree, which produces indistinguishable samples. As the samples increases, manual diagnosis becomes difficult, which decreases the accuracy of fault diagnosis. For accurately and quickly judging the fault type, we propose an improved adaptive activation function and apply it to five types of neural networks. The adaptive activation function improves the negative semi-axis slope of the ReLU activation function by combining the gated channel conversion unit to improve the performance of the deep learning model. The proposed adaptive activation function is compared with the traditional activation function through the fault diagnosis data set and the public data set. The results show that the activation function has better nonlinearity, can improve the generalization performance of deep learning model, the accuracy of fault diagnosis. In addition, the proposed adaptive activation function can also be well embedded in other neural networks.

Near Earth Asteroids represent potential threats to human life because their trajectories may bring them in the proximity of the Earth. Monitoring these objects could help predict future impact events, but such efforts are hindered by the large numbers of objects that pass through the Earth’s vicinity. Additionally, there is also the problem of distinguishing asteroids from other objects in the night sky, which implies sifting through large sets of telescope image data. Within this context, we believe that employing machine learning techniques could greatly improve the detection process by sorting out the most likely asteroid candidates to be reviewed by human experts. At the moment, the use of machine learning techniques is still limited in the field of astronomy and the main goal of the present paper is to study the effectiveness of deep CNNs for the classification of astronomical objects, asteroids in this particular case, by comparing some of the well-known deep convolutional neural networks, including InceptionV3, Xception, InceptionResNetV2 and ResNet152V2. We have applied transfer learning and fine-tuning on these pre-existing deep convolutional networks and from the results that we have obtained one can see the potential of using deep convolutional neural networks in the process of asteroid classification. The InceptionV3 model has the best results in the asteroid class, meaning that by using it, we loose the least number of valid asteroids.

This paper proposes a multimodal fusion 3D target detection algorithm based on the attention mechanism to improve the performance of 3D target detection. The algorithm utilizes point cloud data and information from camera. For image feature extraction, the ResNet50+FPN architecture extracts features at four levels. Point cloud feature extraction employs the voxel method and FCN to extract point and voxel features. The fusion of image and point cloud features is achieved through regional point fusion and regional voxel fusion methods. After information fusion, the Coordinate attention mechanism and SimAM attention mechanism extract fusion features at a deep level. The algorithm's performance is evaluated using the DAIR-V2X dataset. The results show that compared to the Part-A2 algorithm, the proposed algorithm improves the mAP value by 7.9% in BEV view and 7.8% in 3D view at IOU=0.5 (cars) and IOU=0.25 (pedestrians and cyclist). At IOU=0.7 (cars) and IOU=0.5 (pedestrians and cyclist), the mAP value of the SECOND algorithm is improved by 5.4% in the BEV view and 4.3% in the 3D view, compared to other comparison algorithms.

Extracting railway tracks is crucial for creating electronic railway maps. Traditional methods require significant manual labor and resources, while existing neural networks have limitations in efficiency and precision. To address these challenges, a railway track extraction method using an improved DeepLabV3+ model is proposed, which incorporates several key enhancements. Firstly, the encoder part of the method utilizes the lightweight network MobileNetV3 as the backbone extraction network for DeepLabV3+. Secondly, the decoder part adopts the lightweight universal upsampling operator CARAFE for upsampling. Lastly, to address any potential extraction errors, morphological algorithms are applied to optimize the extraction results. Additionally, a dedicated railway track segmentation dataset is created to train and evaluate the proposed method. The experimental results demonstrate that the model achieves impressive performance on the railway track segmentation dataset and DeepGlobe dataset. The MIoU scores are 88.93% and 84.72%, with Recall values of 89.02% and 86.96%. Moreover, the overall accuracy stands at 97.69% and 94.84%. The algorithm's operation time is about 5% lower in comparison to the original network. Furthermore, the morphological algorithm effectively eliminates errors like holes and spots. These findings indicate the model's accuracy, efficiency, and the enhancement brought by the morphological algorithm in error elimination.

The creation of crop-type maps from satellite data has proven challenging, often impeded by a lack of accurate in-situ data. This paper aims to demonstrate a method for crop-type (ie. Maize, Wheat and Other) recognition based on Convolutional Neural Networks using a bottom-up approach. We trained the model with a highly accurate dataset of crowdsourced labelled street-level imagery. Classification results achieved an AUC of 0.87 for wheat, 0.85 for maize and 0.73 for other. Given that wheat and maize are the two most common food crops globally, combined with an ever-increasing amount of available street-level imagery, this approach could help address the need for improved crop-type monitoring globally. Challenges remain in addressing the noisy aspect of street-level imagery (ie. buildings, hedgerows, automobiles, etc.), where a variety of different objects tend to restrict the view and confound the algorithms

Recently, new semantic segmentation and object detection methods have been proposed for direct processing of three-dimensional (3D) LiDAR sensor point clouds. LiDAR can produce highly accurate and detailed 3D maps of natural and man-made environments and is used for sensing in many contexts due to its ability to capture more information and its robustness to dynamic changes in the environment. In addition, the cost of LiDAR sensors has decreased in recent years, which is an important factor for many application scenarios. The challenge with 3D LiDAR sensors like the Velodyne HDL-64E S2 is that they can output large 3D data at up to 1,300,000 points per second, which is difficult to process in real time when applying complex algorithms and models for efficient semantic segmentation. Most existing approaches are either only suitable for relatively small point clouds or rely on computationally intensive sampling techniques to reduce their size. As a result, most of these methods do not work in real-time in realistic field robotics application scenarios, making them unsuitable for practical applications. Systematic point selection is a possible solution to reduce the amount of data to be processed, but although it is memory and computationally efficient, it selects only a small subset of points, which may result in important features being missed. To address this problem, we propose a new approach to semantic segmentation in forestry that uses a systematic sampling method in which the local neighbors of each point are retained to preserve geometric details. Our approach has been shown to process up to 1 million points in a single pass. It outperforms the state of the art in efficient semantic segmentation in large datasets such as Semantic3D. We also present a preliminary study on the performance of LiDAR-only data, i.e., intensity values from LiDAR sensors without RGB values for semi-autonomous robot perception.

The point cloud is a form of three-dimensional data that comprises various detailed features at multiple scales. Due to this characteristic and its irregularity, point cloud analysis based on deep learning is challenging. While previous works utilize the sampling-grouping operation of PointNet++ for feature description and then explore geometry by means of sophisticated feature extractors or deep networks, such operations fail to describe multi-scale features effectively. Additionally, these techniques have led to performance saturation. And it is difficult for standard MLPs to directly "mine" point cloud. To address above problems, we propose the Detail Activation (DA) module, which encodes data based on Fourier transform after sampling and grouping. We activate the channels at different frequency levels from low to high in the DA module to gradually recover finer point cloud details. As training progresses, the proposed Point-MDA can uncover local and global geometries of point cloud progressively. Our experiments show that Point-MDA achieves superior classification accuracy, outperforming PointNet++ by 3.3% and 7.9% in terms of overall accuracy on the ModelNet40 and ScanObjectNN dataset, respectively. Furthermore, it accomplishes this without employing complicated operations, while exploring the full potential of PointNet++.

Once on Mars, maintenance and repair will be crucial for humans as supply chains including Earth andMars will be very complex. Consequently, the raw material available on mars must be processed and used. Factors such as the energy available for material production play just as important a role as the quality of the material that can be produced and the quality of its surface. With the aim of developing and technically implementing a process chain that meets the challenge of producing spare parts from oxygen-reduced Mars regolith, this paper addresses the issue of low-energy handling. Expected statistically distributed high roughnesses of sintered regolith analogs are approximated in this work by parameter variation in the PBF-LB/M process. For low-energy handling, a dry-adhesive microstructure is used. Investigations are carried out to determine the extent to which the rough surface resulting from the manufacturing process can be smoothed by deep rolling in such a way that the microstructure adheres and enables samples to be transported. For the investigated AlSi10Mg samples (12 mm x 12 mm x 10 mm), the surface roughness varies in a wide range from Sa 7.7 µm to Sa 64 µm after the additive manufacturing process, pull-off stresses of up to 6.99 N/cm² could be realized after deep rolling. This represents an increase in pull-off stresses by a factor of 392.94 compared to the pull-off stresses before deep rolling, enabling handling of even larger specimens. It is noteworthy that specimens with roughness values that were previously difficult to handle can be treated post-deep rolling, indicating a potential influence of additional variables that describe roughness or ripples and are associated with the adhesion effect of the microstructure of the dry adhesive.

In the last decade, the genetic engineering world has been shaken up by a relatively new genetic editing tool based on RNA-guided Nucleases (RGNs): the CRISPR/Cas9 system. Since the first report in 1987 and its characterization in 2007 as a bacterial defense mechanism, the interest and research on this system have grown exponentially. CRISPR systems provide immunity to bacteria against invading genetic material; however, with specific modifications in sequence and structure, it becomes a precise editing system that makes it possible to genetically modify almost any organism. There are diverse approaches regarding the refinement of these modifications, such as constructing more accurate nucleases, understanding the cellular context and facing the epigenetic conditions, or re-designing guide RNAs (gRNAs). Considering the critical importance for the correct CRISPR/Cas9 systems performance, our scope will emphasize in the latter approach. Hence, we present an overview of the past and the most recent guide RNA web-based design tools, highlighting their computational architecture and gRNA characteristics evolution through the years. Our study concisely explains the computational approaches that use machine learning techniques, deep neural networks, and large datasets of gRNA/target interactions to make possible both predictions and classifications directed to design, optimize, and create promising gRNAs suitable for future gene therapies.

Deep-ultraviolet (UV) light-emitting diodes (LEDs) based on AlGaN crystals have low light-emission efficiency; therefore, there is a need to improve this light-emission efficiency for a wide range of applications such as water and air sterilizations. UV-light-transparent device structures are considered one of the many solutions toward increasing light output power. To this end, the present study focused on developing a transparent AlGaN-based tunnel junction (TJ) as the anode of a deep-UV LED. Deep-UV LEDs composed of n+/p+-type AlGaN TJs were fabricated under the growth condition that reduced the carrier compensation in the n+-type AlGaN layers. The operating voltage was 10.8 V under the direct current (DC) operation of 63 A/cm2. In addition, magnesium zinc oxide (MgZnO)/Al reflective electrodes were fabricated to enhance the output power of the AlGaN homoepitaxial TJ LED. The output power was 57.3 mW under a DC operation of 63 A/cm2, and it was 1.7 times higher than that realized using the conventional Ti/Al electrodes. The combination of the AlGaN-based TJ and MgZnO/Al reflective contact allows further improvement of the light output power. This study confirms that the AlGaN TJ is a promising UV-transmittance structure that can obtain a high light-extraction efficiency.

Translation of basic animal research to find effective methods of diagnosing and treating human neurological disorders requires parallel analysis infrastructures. Small animals such as mice provide exploratory animal disease models. However, many interventions developed using small animal models fail to translate to human use due to physical or biological differences. Recently, large-animal minipigs have emerged in neuroscience due to both brain similarity and economic advantages. Medical image processing is a crucial part of research as it allows researchers to monitor their experiments and understand disease development. However, although many algorithms are created and optimized for MR analysis of human data, those tools are not directly applicable or sufficiently sensitive to measure minipig data. In this work, we propose PigSNIPE - a pipeline for the automated handling, processing, and analyzing of large-scale data sets of minipig MR images. The pipeline allows for image registration, AC-PC alignment, landmark detection, skull stripping, brainmasks and intracranial volume segmentation (DICE 0.98), tissue segmentation (DICE 0.82), and caudate-putamen brain segmentation (DICE 0.8) in under two minutes. To the best of our knowledge, this is the first automated pipeline tool aimed at large animal images.

According to the World Health Organization, about 15% of the world’s population has some form of disability. Assistive Technology, in this context, contributes directly to the overcoming of difficulties encountered by people with disabilities in their daily lives, allowing them to receive education and become part of the labor market and society in a worthy manner. Assistive Technology has made great advances in its integration with Artificial Intelligence of Things (AIoT) devices. AIoT processes and analyzes the large amount of data generated by IoT devices and applies Artificial Intelligence models, specifically Machine Learning, to discover patterns for generating insights and assisting in decision making. Based on a systematic literature review, this article aims at identifying the Machine Learning models used in multiple different research about Artificial Intelligence of Things applied to Assistive Technology. The survey of the topics approached in this article also highlights the context of such research, their application, IoT devices used, and gaps and opportunities for further development. Survey results show that 50% of the analyzed research address visual impairment, and for this reason, most of the topics cover issues related to computational vision. Portable devices, wearables, and smartphones constituted the majority of IoT devices. Deep Neural Networks represent 81% of the Machine Learning models applied in the reviewed research.

Stellar NGC 6853, also known as the Dumbbell Nebula, is a remarkable example of an elliptical morphological type 1 multiple shell planetary nebula, which is now in one of its last evolutionary stages. This project focuses on examining the distinct varieties of the nebula temperatures and densities using observation data from the Issac Newton Group of Telescopes and comparing them with the observation data from Observatoire de Haute Provence from 14th to 18th of October 2019, with a primary focus on the data from the Issac Newton Telescope (INT). The densities ranged from diffused regions of 376.63 ± 228.97 atoms per cm³ to dense regions of 6441.11 ± 228.90 atoms per cm³; these values are slightly lower approximately by 2000 than the values from the Observatoire de Haute Provence. The radial velocity and the expansion velocity were also calculated using the spectroscopic data and were found to be 29.50 ± 0.38 km⁻¹ and 37.40 ± 0.38 km⁻¹ respectively. This result is expected from a multi-shell nebula that consists of the main body and surrounding halo

Histologically poor differentiation is associated with lymph node metastasis. Thus, pathological evaluation of biopsy specimens is crucial when treating stomach cancers. Deep learning of WSIs is challenging because the images are enormous. Given the computing limitations, patch-level supervised learning methods have been proposed. However, valuable information is lost when dividing WSIs into smaller patches. Another drawback is the need for pixel-level annotation by a pathologist. It is acceptable to differentiate, i.e., grade, gastric cancer at the holistic tissue level (i.e., under low magnification). We developed a weakly supervised learning technique for tissue-level gastric adenocarcinoma histological differentiation (well-to-moderately or poorly differentiated) and applied global reasoning to tissue-level features. The tissue-level AUROCs of the histological differentiation classifiers were 0.953, 0.969, and 0.943, respectively when data from five hospitals were subjected to threefold cross-validation. Comparison of the Grad-CAM heatmaps of the trained classifier and the pathologists’ annotations confirmed that our weakly supervised model exhibited performed well.

A modern definition of “deep sleep” is elusive despite being ubiquitously appreciated as an important physiological state supporting health and homeostasis. In modern times, human deep sleep is identified by specific bioelectric signatures in the electroencephalogram (EEG) emerging somewhere between periods of wakefulness. However, deep sleep has been used to describe states of quiescence well before the first electrical brain recordings in the late 1800s, highlighting its own evolution in both lay and medical literature. Furthermore, EEG states are not only ill-defined in most mammals outside of humans and laboratory rodents, but non-existent in some invertebrates. Given that all organisms rest and do so with seemingly well-defined utility, it remains a challenge linguistically, scientifically, and comparatively define what “deep sleep” means—or what it should—in a research context. Here, I explore standard definitions of deep sleep from a modern, comparative perspective, and discuss potential problems of using a strict and narrow definition of such a fleeting concept that has historically undergone significant updates. Finally, I suggest a path towards resolving inconsistencies around the meaning of “deep sleep” and consider whether it is truly reflected by any one measure.

Recent progress in microfabrication technique allowed the rapid development of neural implants. They are getting categorized as effective tools for clinical practice, especially to treat traumatic and neurodegenerative disorders. Microelectrode arrays already have been used in numerous neural interface devices. Basically, almost all neural implants have been developed based on BCI (Brain Computer Interface) system. When BCI system falls under invasive technique, it is referred as BMI or Brain Machine Interface. BMIs hold promises for neurorehabilitation of motor and sensory function, cognitive state evaluation and treatment of neurological chaos. A directed overview of the field of neural implants is discussed in this article. The aim of this review is to give a brief introduction of neural prosthetics as well as their exciting applications in treating neurological disorders and a deep discussion on their functionality are mentioned. BCI system and their different types, their functionality, their pros and cons, how other neural implants developed, and their present status have been covered. Different possibilities and possible future of deep brain stimulation (DBS), Neuralink, motor and sensory neural prosthetics are further discussed.

The vehicular technology has integrated many features in the system, which enhances the safety and comfort of the user. Among these features, heating, ventilation, and air conditioning (HVAC) is the only feature that maintains the set cabin air temperature (CAT). The user’s command drives the set CAT, and the thermostat provides feedback to the HVAC to maintain the set CAT. The CAT is increased by extracting the heat from the engine surface produced by the fuel combustion, whereas the CAT is reduced by the known processes of the air conditioning system (ACS). Therefore, the CAT driven by the user’s command may not be optimal, and estimating the optimal CAT is still unsolved. In this work, the user was allowed to input a range for CAT instead of a single value. Optimal HVAC criteria were defined, and the CAT was estimated by performing iterative analysis in the user-selected range satisfying the criteria. The HVAC criteria were defined based on two measurable parameters: air conditioning refrigerant fluid pressure (ACRFP) and engine surface temperature (EST) empirically defined as the vector CATOP. In this article, a NARX DL model by mapping the vehicle-level vectors (VLV) to predict the CATOP in real-time using field data obtained from a 2020 Cadillac CT5 test vehicle. Utilising the DL model, CATOP for future time steps were predicted by varying the CAT in the definite range and applying HVAC criteria. Thus, an optimal set CAT was estimated, corresponding to the optimal CATOP defined by the HVAC criteria. We performed the validation of the DL model for multiple datasets using traditional statistical techniques, namely, signal-to-noise ratio (SNR) values, first-order derivatives (FOD), and root-mean-square error (RMSE).

Graphically-rich applications such as games are ubiquitous with attractive visual effects of Graphical User Interface (GUI) that offers a bridge between software applications and end-users. However, various types of graphical glitches may arise from such GUI complexity and have become one of the main component of software compatibility issues. Our study on bug reports from game development teams in NetEase Inc. indicates that graphical glitches frequently occur during the GUI rendering and severely degrade the quality of graphically-rich applications such as video games. Existing automated testing techniques for such applications focus mainly on generating various GUI test sequences and check whether the test sequences can cause crashes. These techniques require constant human attention to captures non-crashing bugs such as bugs causing graphical glitches. In this paper, we present the first step in automating the test oracle for detecting non-crashing bugs in graphically-rich applications. Specifically, we propose GLIB based on a code-based data augmentation technique to detect game GUI glitches. We perform an evaluation of GLIB on 20 real-world game apps (with bug reports available) and the result shows that GLIB can achieve 100\% precision and 99.5\% recall in detecting non-crashing bugs such as game GUI glitches. Practical application of GLIB on another 14 real-world games (without bug reports) further demonstrates that GLIB can effectively uncover GUI glitches, with 48 of 53 bugs reported by GLIB having been confirmed and fixed so far.

Recommender systems have been applied in a wide range of domains such as e-commerce, media, banking, and utilities. This kind of system provides personalized suggestions based on large amounts of data in order to increase user satisfaction. These suggestions help client select products, while organizations can increase the consumption of a product. In the case of social data, sentiment analysis can help gain better understanding of a user’s attitudes, opinions and emotions, which is beneficial to integrate in recommender systems for achieving higher recommendation reliability. On the one hand, this information can be used to complement explicit ratings given to products by users. On the other hand, sentiment analysis of items that can be derived from online news services, blogs, social media or even from the recommender systems themselves is seen as capable of providing better recommendations to users. In this study, we present and evaluate a recommendation approach that integrates sentiment analysis into collaborative filtering methods. The recommender system proposal is based on an adaptive architecture, which includes improved techniques for feature extraction and deep learning models based on sentiment analysis. The results of the empirical study performed with two popular datasets show that sentiment–based deep learning models and collaborative filtering methods can significantly improve the recommender system’s performance.

This work aims to contribute to better understanding the use of public street spaces. (1) Background: In this sense, with a multidisciplinary approach, the objective of this work is to propose an experimental and reproducible method on a large scale. (2) Study area: The applied methodology uses artificial intelligence to analyze Google Street View (GSV) images at street level. (3) Method: The purpose is to validate a methodology that allows to characterize and quantify the use (pedestrians and cars) of some squares in Rome belonging to different historical periods. (4) Results: Through the use of machine vision techniques, typical of artificial intelligence and which use convolutional neural networks, a historical reading of some selected squares is proposed with the aim of interpreting the dynamics of use and identifying some critical issues in progress. (5) Conclusions: This work validated the usefulness of a method applied to the use of artificial intelligence for the analysis of GSV images at street level.

Segmentation of Magnetic Resonance Images (MRI) of abdominal organs is useful for analysis prior to surgical procedures and for further processing. Deep Learning (DL) has become the standard, researchers have proposed improvements that include multiple views, ensembles and voting. Loss function alternatives, while being crucial to guide automated learning, have not been compared in detail. In this work we analyze limitations of popular metrics and their use as loss, study alternative loss variations based on those and other modifications and search for the best approach. An experimental setup was necessary to assess the alternatives. Results for the top scoring network and top scoring loss show improvements between 2 and 11 percentage points (pp) in Jaccard Index (JI), depending on organ and patient (sequence), for a total of 22 pp over 4 organs, all this being obtained just by choosing the best performing loss function instead of cross-entropy or dice. Our results apply directly to MRI of abdominal organs, with important practical implications for other architectures, as they can be applied easily to any of them. They also show the worth of variants of loss function and loss tuning, with future work needed to generalize and test in other contexts.

Recently deep cross-modal hashing networks have received increasing interests due to its superior query efficiency and low storage cost. However, most of existing methods concentrate less on hash representations learning part, which means the semantic information of data cannot be fully used. Furthermore, they may neglect the high-ranking relevance and consistency of hash codes. To solve these problems, we propose a Self-Attention and Adversary Guided Hashing Network (SAAGHN). Specifically, it employs self-attention mechanism in hash representations learning part to extract rich semantic relevance information. Meanwhile, in order to keep invariability of hash codes, adversarial learning is adopted in the hash codes learning part. In addition, to generate higher-ranking hash codes and avoid local minima early, a new batch semi-hard cosine triplet loss and a cosine quantization loss are proposed. Extensive experiments on two benchmark datasets have shown that SAAGHN outperforms other baselines and achieves the state-of-the-art performance.

We present a method to improve the reconstruction and generation performance of variational autoencoder (VAE) by injecting an adversarial learning. On the other hand, instead of comparing the reconstructed with the original data to calculate the reconstruction loss, we use a consistency principle for deep features. The training process of the VAE is then divided into two steps, training the encoder and then training the decoder. By using this two-step learning process, our method can be more widely used in applications other than image processing. While training the encoder, the label information is integrated to better structure the latent space in a supervised way. The adversarial constraints allow the decoder to generate data with better authenticity and more realistic than the conventional VAE. We present experimental results to show that our method gives better performance than the original VAE.

Significant progress has made in pattern recognition technology. However, one obstacle that has not yet overcome is the recognition of words in the Brahmi script, specifically the identification of characters, compound characters, and word. This study proposes the use of the deep convolutional neural network with dropout to recognize the Brahmi words. This study also proposed a DCNN for Brahmi word recognition and a series of experiments are performed on standard Brahmi dataset. The practical operation of this method was systematically tested on accessible Brahmi image database, achieving 92.47% recognition rate by CNN with dropout respectively which is among the best while comparing with the ones reported in the literature for the same task.

In this paper we address the understanding of the problem, of why a deep learning model decides that an individual is eligible for a loan or not. Here we propose a novel approach for inferring, which attributes matter the most, for making a decision in each specific individual case. Specifically we leverage concepts from neural attention to devise a novel feature wise attention mechanism. As we show, using real world datasets, our approach offers unique insights into the importance of various features, by producing a decision explanation for each specific loan case. At the same time, we observe that our novel mechanism, generates decisions which are much closer to the decisions generated by human experts, compared to the existent competitors.

The latest methods based on deep learning have achieved amazing results regarding the complex work of inpainting large missing areas in an image. This type of method generally attempts to generate one single "optimal" inpainting result, ignoring many other plausible results. However, considering the uncertainty of the inpainting task, one sole result can hardly be regarded as a desired regeneration of the missing area. In view of this weakness, which is related to the design of the previous algorithms, we propose a novel deep generative model equipped with a brand new style extractor which can extract the style noise (a latent vector) from the ground truth image. Once obtained, the extracted style noise and the ground truth image are both input into the generator. We also craft a consistency loss that guides the generated image to approximate the ground truth. Meanwhile, the same extractor captures the style noise from the generated image, which is forced to approach the input noise according to the consistency loss. After iterations, our generator is able to learn the styles corresponding to multiple sets of noise. The proposed model can generate a (sufficiently large) number of inpainting results consistent with the context semantics of the image. Moreover, we check the effectiveness of our model on three databases, i.e., CelebA, Agricultural Disease, and MauFlex. Compared to state-of-the-art inpainting methods, this model is able to offer desirable inpainting results with both a better quality and higher diversity. The code and model will be made available on https://github.com/vivitsai/SEGAN.

Fusarium head blight (FHB) disease is extensively distributed worldwide. This disease damages grain quality and reduces yield. The detection of this disease in a high throughput way is crucial to planters and breeders. Our study focused on developing a method for processing wheat color images and accurately detecting disease areas using deep learning and image processing techniques. The color images of wheat at the milky stage were collected and processed to construct datasets, which were used to retrain a deep convolutional neural network model using transfer learning. Testing results showed that the model can detect spikes, and the coefficient of determination of the number of spikes between the manual count and the detection was 0.80. The model was assessed, and the mean average precision for the testing dataset was 0.9201. On the basis of the results of spike detection, a new color feature was applied to obtain the gray image of each spike. Then, a modified region growing algorithm was implemented to segment and detect the diseased areas of each spike. Results show that the region growing algorithm performs better than K-means and Otsu’s method in segmenting the FHB disease. Overall, this study demonstrates that deep learning techniques enable the accurate detection of FHB in wheat using color images, and the proposed method can effectively detect spikes and diseased areas, thereby improving the efficiency of FHB detection.

Complementary Metal-Oxide-Semiconductor (CMOS) is a typical image sensor that has a wide range of applications. However, considering the limitations of the weather condition and hardware cost, it is hard to capture high-resolution images by CMOS sensor. Recently, Super-Resolution (SR) techniques for image restoration has been gaining attentions due to its excellent performance. Under the powerful learning ability, Generative Adversarial Networks (GANs) have been proved to achieve great success. In this paper, we propose the Advanced Generative Adversarial Networks (AGAN) to efficiently correct these issues; 1) we design a Laplacian pyramid framework as pre-trained module, which is beneficial to provide multi-scale features for our input. 2) at each feature block, a convolutional skip-connections network, which may contain some latent information, is significant for generative model to reconstruct a plausible-looking image; 3) considering that edge details usually play an important role in image generation, a novel perceptual loss function is defined to train and seek optimal parameters. It is effective to achieve excellent and compelling quality captured by CMOS sensor. Quantitative and qualitative evaluations have been demonstrated that our algorithm not only fully takes advantage of Convolutional Neural Networks (CNNs) to improve the image quality, but also performs better than previous GAN algorithms for super-resolution task.

Geologists employ high-spatial-resolution (HR) remote sensing (RS) data for many diverse applications as they effectively reflect detailed geological information, enabling high-quality and efficient geological surveys. Applications of HR RS data to geological and related fields have grown recently. By analyzing these applications, we can better understand the results of previous studies and more effectively use the latest data and methods to efficiently extract key geological information. HR optical remote sensing data are widely used in geological hazard assessment, seismic monitoring, mineral exploitation, glacier monitoring, and mineral information extraction due to high accuracy and clear object features. Compared with optical satellite images, synthetic-aperture radar (SAR) images are stereoscopic and exhibit clear relief, strong performance, and good detection of terrain, landforms, and other information. SAR images have been applied to seismic mechanism research, volcanic monitoring, topographic deformation, and fault analysis. Furthermore, a multi-standard maturity analysis of the geological applications of HR images using literature from the Science Citation Index reveals that optical remote sensing data are superior to radar data for mining, geological disaster, lithologic, and volcanic applications, but inferior for earthquake, glacial, and fault applications. Therefore, geological remote sensing research needs to be truly multidisciplinary or interdisciplinary, ensuring more detailed and efficient surveys through cross-linking with other disciplines. Moreover, the recent application of deep learning technology to remote sensing data extraction has improved automatic processing and data analysis capabilities.

Rockburst is a hazardous phenomenon in deep tunnels influenced by geological structural planes like faults, joints, and shear planes. Small-scale shear-plane-like structures have damaging impact on the boundaries of the tunnel, which accumulate high stresses. Such a shear plane was exposed in the side wall of the right headrace tunnel in the Neelum-Jehlum Hydropower Project. This project is constructed in the tectonically active Himalayas with high stress conditions. A shear plane combined with high stress conditions is very dangerous in deep excavations. The influence of a shear zone on rockburst occurrence near a tunnel is studied. The FLAC3D explicit code simulated the shear zone in the right tunnel, revealing that the stresses are concentrated near the shear zone, with no stress concentration in the left tunnel, which has neither shear zone nor rockburst. Rock mass was displaced near this shear zone even before excavation. Modeling results confirm that the side wall shear zone of the tunnel has a major influence on rockburst occurrence. A shear slip along this plane released huge amounts of energy, causing human fatalities and property damage. A numerical simulation validates the actual conditions and helps us understand the phenomenon of stress concentration near the shear zone and its impact during deep tunneling.

Oil spillage over a sea or ocean’s surface is a threat to marine and coastal ecosystems. Spaceborne synthetic aperture radar (SAR) data has been used efficiently for the detection of oil spills due to its operational capability in all-day all-weather conditions. The problem is often modeled as a semantic segmentation task. The images need to be segmented into multiple regions of interest such as sea surface, oil spill, look-alikes, ships and land. Training of a classifier for this task is particularly challenging since there is an inherent class imbalance. In this work, we train a convolutional neural network (CNN) with multiple feature extractors for pixel-wise classification; and introduce to use a new loss function, namely ‘gradient profile’ (GP) loss, which is in fact the constituent of the more generic Spatial Profile loss proposed for image translation problems. For the purpose of training, testing and performance evaluation, we use a publicly available dataset with selected oil spill events verified by the European Maritime Safety Agency (EMSA). The results obtained show that the proposed CNN trained with a combination of GP, Jaccard and focal loss functions can detect oil spills with an intersection over union (IoU) value of 63.95%. The IoU value for sea surface, look-alikes, ships and land class is 96.00%, 60.87%, 74.61% and 96.80%, respectively. The mean intersection over union (mIoU) value for all the classes is 78.45%, which accounts for a 13% improvement over the state of the art for this dataset. Moreover, we provide extensive ablation on different Convolutional Neural Networks (CNNs) and Vision Transformers (ViTs) based hybrid models to demonstrate the effectiveness of adding GP loss as an additional loss function for training. Results show that GP loss significantly improves the mIoU and F1 scores for CNNs as well as ViTs based hybrid models. GP loss turns out to be a promising loss function in the context of deep learning with SAR images.

The Bayes filters, such as Kalman and particle filters, have been used in sensor fusion to integrate two sources of information and obtain the best estimate of the unknowns. Efficient integration of multiple sensors requires deep knowledge of their error sources and it is not trivial for complicated sensors, such as Inertial Measurement Unit (IMU). Therefore, IMU error modelling and efficient integration of IMU and Global Navigation Satellite System (GNSS) observations has remained a challenge. In this paper, we develop deep Kalman filter to model and remove IMU errors and consequently, improve the accuracy of IMU positioning. In other words, we add modelling step to the prediction and update steps of Kalman filter and the IMU error model is learned during integration. Therefore, our deep Kalman filter outperforms Kalman filter and reaches higher accuracy.

In any document, graphical elements like tables, figures, and formulas contain essential information. The processing and interpretation of such information require specialized algorithms. Off-the-shelf OCR components cannot process this information reliably. Therefore, an essential step in document analysis pipelines is to detect these graphical components. It leads to a high-level conceptual understanding of the documents that makes digitization of documents viable. Since the advent of deep learning, the performance of deep learning-based object detection has improved many folds. In this work, we outline and summarize the deep learning approaches for detecting graphical page objects in the document images. Therefore, we discuss the most relevant deep learning-based approaches and state-of-the-art graphical page object detection in document images. This work provides a comprehensive understanding of the current state-of-the-art and related challenges. Furthermore, we discuss leading datasets along with the quantitative evaluation. Moreover, it discusses briefly the promising directions that can be utilized for further improvements.

Periodontics is a complex field characterized by a constantly growing body of research, posing a challenge for researchers and stakeholders striving to stay abreast of its evolving literature. Tradi-tional bibliometric surveys, while accurate, are labor-intensive and not scalable to meet the demands of such rapidly expanding domains. The aim of this study was to explore the uses of a top-ic-modeling algorithm to investigate the structure of the periodontics research field. To this purpose, we employed BERTopic to automatically sift through a substantial corpus of periodontics manu-scripts published in MEDLINE from 2009 to 2024 and identify topics using two sets of settings, to get a broader and high detailed view of the field, respectively. BERTopic identified and categorized 31 topics, unveiling the internal structure and the relations between publication macroareas. When we used the high granularity settings, we identified 2428 topics, which revealed hotspots within the field. Specific regression highlighted novel research areas such as artificial intelligence and exo-somes. In conclusion, BERTopic offers a flexible tool to provide insights into periodontics, its most recent trends, and serve as a useful instrument to approach the field and direct future research.

Introduction: Chronic ulcers significantly burden healthcare systems, requiring precise measurement and assessment for effective treatment. Traditional methods, such as manual segmentation, are time-consuming and error-prone. This review evaluates the potential of artificial intelligence (AI)--powered mobile apps for automated ulcer segmentation and their application in clinical settings. Methods: A comprehensive literature search was conducted across PubMed, CINAHL, Cochrane and Google Scholas databases. The review focused on mobile apps that use fully automatic AI algorithms for wound segmentation. Apps requiring additional hardware or needing more technical documentation were excluded. Vital technological features, clinical validation, and usability were analysed. Results: Ten mobile apps were identified, showing varying levels of segmentation accuracy and clinical validation. However, many apps did not publish sufficient information on the segmentation methods or algorithms used, and most lacked details on the databases employed for training their AI models. Additionally, several apps were unavailable in public repositories, limiting their accessibility and independent evaluation. These factors challenge their integration into clinical practice despite promising preliminary results. Discussion: AI-powered mobile apps offer significant potential for improving wound care by enhancing diagnostic accuracy and reducing the burden on healthcare professionals. Nonetheless, the lack of transparency regarding segmentation techniques, unpublished databases, and limited availability of many apps in public repositories remain substantial barriers to widespread clinical adoption. Conclusion: AI-driven mobile apps for ulcer segmentation could revolutionise chronic wound management. However, overcoming limitations related to transparency, data availability, and accessibility is essential for their successful integration into healthcare systems.

Due to the influence of factors such as user consumption habits and manufacturer service quality, e-commerce platforms cannot push personalized product information to any consumer. This project studies the push of product information on e-commerce platforms based on deep neural networks. It mines the data of users' online shopping behaviors such as browsing, collecting, and adding to cart on e-commerce platforms, and pre-processes the mined user behavior data by cleaning, integrating, and normalizing them. It builds a deep bidirectional Transformer model to learn users' historical behavior data and predict the most probable products that meet users' behavior needs, thereby realizing the automatic push of product information on e-commerce platforms. The experimental results show that the F1 value of the push result of product information on e-commerce platforms under the method designed by this algorithm is 0.97, which confirms the effectiveness and superiority of this method.

Two-dimensional ultrasound (2D US) is commonly used in clinical settings for its cost-effectiveness and non-invasiveness, but it is limited by spatial orientation and operator dependency. Three-dimensional ultrasound (3D US) overcomes these limitations by adding a third dimension and enhancing integration with other imaging modalities. Advances in deep learning (DL) have further propelled the viability of freehand image-based 3D reconstruction, broadening clinical applications in intraoperative and point-of-care (POC) settings. This review evaluates state-of-the-art freehand 3D US reconstruction methods that eliminate the need for external tracking devices, focusing on experimental setups, data acquisition strategies, and reconstruction methodologies. PubMed, Scopus, and IEEE Xplore were searched for studies since 2014 following PRISMA guidelines, excluding those using additional imaging or tracking systems other than inertial measurement units (IMUs). Fourteen eligible studies were analyzed, showing a shift from traditional speckle decorrelation towards DL-based methods, particularly Convolutional Neural Networks (CNNs). Variability in datasets and evaluation methods hindered a comprehensive quantitative comparison, but notable accuracy improvements were observed with IMUs and integration of contextual and temporal information within CNNs. These advancements enhance freehand 3D US reconstruction feasibility, though variability limits definitive conclusions about the most effective methods. Future research should focus on improving precision in complex trajectories and adaptability across clinical scenarios.

The objective of image super-resolution is to reconstruct a high-resolution (HR) image with the prior knowledge from one or several low-resolution (LR) images. However, in the real world, due to the limited complementary information, the performance of both single-frame and multi-frame super-resolution reconstruction degrades rapidly as the magnification increases. In this paper, we propose a novel two-step image super resolution method concatenating multi-frame super-resolution (MFSR) with single-frame super-resolution (SFSR), to progressively upsample images to the desired resolution. The proposed method consisting of an L0-norm constrained reconstruction scheme and an enhanced residual back-projection network, integrating the flexibility of the variational model-based method and the feature learning capacity of the deep learning-based method. To verify the effectiveness of the proposed algorithm, extensive experiments with both simulated and real world sequences were implemented. The experimental results show that the proposed method yields superior performance in both objective and perceptual quality measurements, compared to the baseline super-resolution algorithms in the cascade model. In addition, the experiment indicates that this cascade model can be robustly applied to different SFSR and MFSR methods.

In addressing the multifaceted problem of multiple-input multiple-output (MIMO) detection in wireless communication systems, this work highlights the pressing need for enhanced detection reliability under variable channel conditions and MIMO antenna configurations. We propose a novel method that sets a new standard for deep unfolding approaches to MIMO detection by integrating the iterative conjugate gradient method with Tikhonov regularization, combining the adaptability of modern deep learning techniques with the robustness of classical regularization. Unlike conventional techniques, our strategy treats the regularization parameter of Tikhonov regularization as well as step size values and search direction coefficients of conjugate gradient (CG) method as trainable parameters within the deep learning framework, allowing for dynamic modification according to channel conditions and MIMO antenna configurations. Detection performance is significantly improved by the proposed approach in variety of conditions. In different MIMO settings, the suggested method consistently shows better bit error rate (BER) and normalized minimum mean square error (NMSE) performance. Across a range of MIMO configurations and channel conditions, the proposed method exhibits significantly lower BER and NMSE values than well-known techniques such as DetNet and CG. The proposed method has superior performance over CG and other model-oriented methods, especially in small number of iterations. Consequently, the simulation results demonstrate the flexibility of the proposed approach, making it a viable choice for MIMO systems with a range of antenna configurations and different channnel conditions.

The exponential growth of multimodal content across social media platforms, comprising text, images, audio, and video, has catalyzed substantial interest in artificial intelligence, particularly in multi-modal sentiment analysis (MSA). This study presents a comprehensive survey of 30 research papers published between 2020 and 2024 by eminent publishers such as Elsevier, ACM, IEEE, Springer, and others indexed in Google Scholar. Our analysis primarily focuses on exploring multimodal fusion techniques and features, with specific emphasis on the integration of text and image data. Additionally, the article offers an overview of the evolution, definition, and historical context of MSA. It delves into the current challenges and potential advantages of MSA, investigating recent datasets and sophisticated models. Furthermore, the study provides insights into prospective research directions. Notably, this review offers valuable recommendations for advancing research and developing more robust MSA models, thus serving as a valuable resource for both academic and industry researchers engaged in this burgeoning field.

Deep hypothermic circulatory arrest (DHCA) is safe, but subtle neurodevelopmental deficits may persist far beyond the perioperative period. We investigated the relationship between DHCA duration and neurodevelopmental outcomes in young children undergoing cardiac surgery with DHCA. Children aged < 42 months, including neonates who underwent cardiac surgery using DHCA without regional perfusion techniques, were included as the DHCA group. Children in the same age range undergoing cardiac surgery without DHCA were included as the control group. All patients underwent neurodevelopmental assessment using the Bayley Scales of Infant and Toddler Development (BSTID) by a trained pediatrician. Seventeen of 25 DHCA and six of 10 control patients completed the BSTID assessment. The groups showed no preoperative, operative, or postoperative significant differences. In adjusted multivariable analysis, prematurity and age at assessment were significant changing predictors of each of the BSTID components (p < 0.001), other than the gross motor component, where only age at assessment was a significant adjusting predictor. Longer DHCA was associated with lower fine and gross motor BSTID components; however, the association was not statistically significant (p = 0.06). Use of DHCA without regional perfusion techniques may be associated with less optimal neurodevelopmental outcomes if DHCA has a long duration.

In order to shorten the detection time and improve the average precision on embedded devices, A lightweight and high accuracy model is proposed for passion fruit in complex environments (backlight, occlusion, overlap, sunny, cloudy, rainy). Firstly, replacing the backbone network of YOLOv5 with a lightweight GhostNet model reduces the number of parameters and computation while improving detection speed. Secondly, a new feature branch is added to the GhostNet network, and the feature fusion layer in the neck network is reconstructed to effectively combine the lower-level and higher-level features, which not only improves the accuracy of the model but also maintains its lightweight. Finally, the knowledge distillation methods are used to transfer the knowledge from the more capable teacher model to the less capable student model, which significantly improving the detection accuracy. The improved model is denoted as G-YOLO-NK. The average accuracy of the G-YOLO-NK network is 96.00%, which is 1.00% higher than the original YOLOv5s model. Furthermore, the improved model size is 7.14MB, reduced to half of the original model, and the real-time detection frame rate is 11.25 FPS on the Jetson Nano. Compared to the state-of-the-art model, the proposed model outperforms them in terms of average precision and detection performance. The present work provides an effective model for real-time detection of passion fruits in complex orchard scenes, which can provide valuable technical support for the development of orchard picking robots and greatly improve the intelligence level of orchards.