Estimating the remaining useful life (RUL) of components is a crucial task to enhance the reliability, safety, productivity, and to reduce maintenance cost. In general, predicting the RUL of a component includes constructing a health indicator ( ) to infer the current condition of the component, and modelling the degradation process, to estimate the future behavior. Although many signal processing and data-driven based methods were proposed to construct the , most of the existing methods are based on manual feature extraction techniques, and need the prior knowledge of experts, or rely on a large amount of failure data. Therefore, in this study, a new data-driven method based on the convolutional autoencoder (CAE) is presented to construct the . For this purpose, the continuous wavelet transform (CWT) technique is used to convert the raw acquired vibrational signals into a two-dimensional image; then, the CAE model is trained by the healthy operation dataset. Finally, the Mahalanobis Distance (MD) between the healthy and failure stages is measured as the . The proposed method is tested on a benchmark bearing dataset and compared with several other traditional construction models. Experimental results indicate that the constructed exhibits a monotonically increasing degradation trend and has a good performance to detect incipient faults.

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.

Autoencoders are neural networks that have applications in denoising processes. Their use is widely reported in imaging (2D), though 1D series can also benefit from this function. Here, three canonical waveforms are used to train a neural network and achieve a signal-to-noise reduction with curves whose noise energy was above that of the signals. A real-world test is carried out with the same autoencoder subjected to a set of time series corrupted by noise generated by a zener diode, biased on the avalanche region. Results showed that, observed some guidelines, the autoencoder can indeed denoise 1D waveforms usually observed in electronics, particularly square waves found in digital circuits.

The term "code smell" refers to an indication of a problem with the quality of source code. Numerous studies have been conducted to identify problematic features in source code. Initially, the focus was on utilizing metric-based and heuristic-based approaches. In recent years, however, there has been a shift towards using machine learning and deep learning (DL) techniques for smell detection. Nevertheless, the current algorithms are still considered to be in the early stages of development. Recognizing the challenges associated with identifying smells using DL methods, both academics and software developers have made efforts to address these obstacles. This work involves constructing and evaluating new DL models for code smell detection. Two models are built upon stacked autoencoders, employing a hybrid architecture that combines bidirectional long short-term memory and convolutional neural network components.

Self-supervised learning is a method that learns general representation from unlabeled data. Masked image modeling (MIM), one of the generative self-supervised learning methods, has drawn attention showing state-of-the-art performance on various downstream tasks, though showing poor linear separability resulting from the token-level approach. In this paper, we propose a contrastive learning-based multi-view masked autoencoder for MIM, exploiting an image-level approach by learning common features from two different augmented views. We strengthen MIM by learning long-range global patterns from contrastive loss. Our framework adopts simple encoder-decoder architecture, learning rich and general representation by following a simple process: 1) two different views are generated from an input image with random masking and by contrastive loss, we can learn semantic distance of the representations generated by an encoder. By applying a high mask ratio, 80%, it works as strong augmentation and alleviates the representation collapse problem. 2) With reconstruction loss, decoder learns to reconstruct an original image from the masked image. We assess our framework by several experiments on benchmark datasets of image classification, object detection, and semantic segmentation. We achieve 84.3% fine-tuning accuracy on ImageNet-1K classification and 76.7% in linear probing, exceeding previous studies and show promising results on other downstream tasks. Experimental results demonstrate that our work can learn rich and general image representation by applying contrastive loss to masked image modeling.

The task of human action recognition (HAR) based on skeleton data is a challenging yet crucial technique owing to its wide-ranging applications in numerous domains, including patient monitoring, security surveillance, and observation of human-machine interactions. While numerous algorithms have been proposed in an attempt to distinguish between a myriad of activities, most practical applications necessitate highly accurate detection of specific activity types. This study proposes a novel and highly accurate spatiotemporal graph autoencoder network for HAR based on skeleton data. Furthermore, an extensive investigation was conducted employing diverse modalities. To this end, a spatiotemporal graph autoencoder was constructed to automatically learn both spatial and temporal patterns from human skeleton datasets. The powerful graph convolutional network, designated as GA-GCN, developed in this study, notably outperforms the majority of existing state-of-the-art methods when evaluated on two common datasets, namely NTU RGB+D and NTU RGB+D 120. On the first dataset, the proposed approach achieved accuracies of 92.3\% and 96.8\% for the cross-subject and cross-view evaluations, respectively. On the more challenging NTU RGB+D 120 dataset, GA-GCN attained accuracies of 88.8\% and 90.4\% for the cross-subject and cross-set evaluations, respectively.

Speech Emotion Recognition (SER) is an interesting and difficult problem to handle. In this paper, we deal with it through the implementation of deep learning networks. We have designed and implemented six different deep learning networks, a Deep Belief Network (DBN), a simple deep neural network (SDNN), a LSTM network (LSTM), a LSTM network with the addition of an attention mechanism (LSTM-ATN), a Convolutional neural network (CNN), and a Convolutional neural network with the addition of an attention mechanism (CNN-ATN), having in mind, apart from solving the SER problem, to test the impact of attention mechanism to the results. Dropout and Batch Normalization techniques are also used to improve the generalization ability (prevention of overfitting) of the models as well as to speed up the training process. The Surrey Audio-Visual Expressed Emotion database (SAVEE), and the Ryerson Audio-Visual Database (RAVDESS) database were used for training and evaluation of our models. The results showed that networks with the addition of the attention mechanism did better than the others. Furthermore, they showed that CNN-ATN was the best among tested networks, achieving an accuracy of 74% for the SAVEE and 77% for the RAVDESS dataset, and exceeded existing state-of-the-art systems for the same datasets.

Many works have recently identified the need to combine deep learning with extreme learning to strike a performance balance with accuracy especially in the domain of multimedia applications. Considering this new paradigm, namely convolutional extreme learning machine (CELM), we present a systematic review that investigates alternative deep learning architectures that use extreme learning machine (ELM) for a faster training to solve problems based on image analysis. We detail each of the architectures found in the literature, application scenarios, benchmark datasets, main results, advantages, and present the open challenges for CELM. We follow a well structured methodology and establish relevant research questions that guide our findings. We hope that the observation and classification of such works can leverage the CELM research area providing a good starting point to cope with some of the current problems in the image-based computer vision analysis.

Performing predictive maintenance (PdM) is challenging for many reasons. Dealing with large datasets which may not contain run-to-failure data (R2F) complicates PdM even more. When no R2F data are available, identifying condition indicators (CIs), estimating the health index (HI), and thereafter, calculating a degradation model for predicting the remaining useful lifetime (RUL) are merely impossible using supervised learning. In this paper, a 3 dof delta robot used for pick and place task is studied. In the proposed method, autoencoders (AEs) are used to predict when maintenance is required based on the signal sequence distribution and anomaly detection, which is vital when no R2F data is available. Due to the sequential nature of the data, non-linearity of the system, and correlations between parameter time series, convolutional layers are used for feature extraction. Thereafter, a sigmoid function is used to predict the probability of having an anomaly given CIs acquired from AEs. This function can be manually tuned given the sensitivity of the system or optimized by solving a minimax problem. Moreover, the proposed architecture can be used for fault localization for the specified system. Additionally, the proposed method is capable of calculating RUL using Gaussian process (GP), as a degradation model, given HI values as its input.

An adaptive deep neural network is used in an inverse system identification setting to approximate the inverse of a nonlinear plant with the aim of constituting the plant controller by copying to the latter the weights and architecture of the converging deep neural network. This deep learning (DL) approach to the adaptive inverse control (AIC) problem is shown to outperform adaptive filtering techniques and algorithms normally used in adaptive control, especially when the plant is nonlinear. The deeper the controller the better the inverse function approximation, provided that the nonlinear plant have an inverse and that this inverse can be approximated. Simulation results prove the feasibility of this DL-based adaptive inverse control scheme. The DL-based AIC system is robust to parameter change of the nonlinear plant in that, under such change, the plant output reassumes the value of the reference signal considerably faster than with the adaptive filter counterpart of the deep neural network. Settling times and rise times of the step response are shown to improve in the DL-based AIC system.

The rising number of unknown-type attacks on the Internet emphasizes the significance of developing efficient intrusion detection systems, even if machine learning-based techniques can detect unknown types of attacks. The necessity for innovative techniques is highlighted by the possibility that traditional machine learning techniques will not be sufficient for identifying these unknown types of attacks. In this research, we address this difficulty by proposing a deep learning-based solution: the log-cosh variational autoencoder (LVAE). When it comes to understanding intricate data distributions and creating freshly reconstructed data, LVAE inherits the strong modeling skills of the variational autoencoder (VAE). To better imitate discrete features of actual attacks and generate unknown types of attacks, this study develops an effective reconstruction loss term employing the logarithmic hyperbolic cosine (log-cosh) function in the log-cosh variational autoencoder (LVAE). When compared to conventional VAEs, LVAE exhibits promising potential for effectively generating data that closely resembles an unknown attack, a crucial capability for increasing the unknown attack detection rate. To categorize the generated unknown-type data, eight feature extraction and classification techniques were used. Using the most recent CICIDS2017 dataset, numerous experiments were carried out, training under varying amounts of real and unknown-type attacks. Our optimal experimental results outperformed several state-of-the-art techniques with accuracy and average F1 scores of 99.89% and 99.83%, respectively. Outstanding results were also shown by the suggested LVAE strategy for producing unknown attack data. In general, our work sets a strong basis for the accurate and efficient identification of unknown types of attacks and contributes to the development of intrusion detection techniques.

In industrial settings, gears play a crucial role by assisting various machinery functions such as speed control, torque manipulation, and altering motion direction.The malfunction or failure of these gear components can have serious repercussions, resulting in production halts and financial losses. As a result, there is an increasing requirement to monitor the state of these components in order to avoid such issues from occurring. To address this need, research efforts have focused on early defect detection in gears in order to reduce the impact of possible failures. This study focused on analyzing vibration and thermal datasets from two extruder machine gearboxes using an autoencoder Long Short-Term Memory (LSTM) model. The major goal is to implement an outlier detection approach to detect and classify defects. The results of this study highlighted the extraordinary performance of the Autoencoder LSTM model, which achieved an impressive accuracy rate of 94.42% in recognizing malfunctioning gearboxes within the extruder machine system. Furthermore, the study used a thorough global metrics evaluation methodology to further test the model’s dependability and efficacy, consequently substantiating the proposed approach’s validity.

The COVID-19 pandemic situation has created even more difficulties in the quick identification and screening of the COVID-19 patients for the medical specialists. Therefore, a significant study is necessary for detecting COVID-19 cases using an automated diagnosis method, which can aid in controlling the spreading of the virus. In this paper, the study suggests a Deep Convolutional Neural Network-based multi-classification approach (COV-MCNet) using eight different pre-trained architectures such as VGG16, VGG19, ResNet50V2, DenseNet201, InceptionV3, MobileNet, InceptionResNetV2, Xception which are trained and tested on the X-ray images of COVID-19, Normal, Viral Pneumonia, and Bacterial Pneumonia. The results from 3-class (Normal vs. COVID-19 vs. Viral Pneumonia) showed that only the ResNet50V2 model provides the highest classification performance (accuracy: 95.83%, precision: 96.12%, recall: 96.11%, F1-score: 96.11%, specificity: 97.84%) compared to rest of the models. The results from 4-class (Normal vs. COVID-19 vs. Viral Pneumonia vs. Bacterial Pneumonia) demonstrated that the pre-trained model DenseNet201 provides the highest classification performance (accuracy: 92.54%, precision: 93.05%, recall: 92.81%, F1-score: 92.83%, specificity: 97.47%). Notably, the ResNet50V2 (3-class) and DenseNet201 (4-class) models in the proposed COV-MCNet framework showed higher accuracy compared to the rest six models. This indicates that the designed system can produce promising results to detect the COVID-19 cases on the availability of more data. The proposed multi-classification network (COV-MCNet) significantly speeds up the existing radiology-based method, which will be helpful to the medical community and clinical specialists for early diagnosis of the COVID-19 cases during this pandemic.

Autonomous image recognition has numerous potential applications in the field of planetary science and geology. For instance, having the ability to classify images of rocks would allow geologists to have immediate feedback without having to bring back samples to the laboratory. Also, planetary rovers could classify rocks in remote places and even in other planets without needing human intervention. Shu et al. classified 9 different types of rock images using a Support Vector Machine (SVM) with the image features extracted autonomously. Through this method, the authors achieved a test accuracy of 96.71%. In this research, Convolutional Neural Networks(CNN) have been used to classify the same set of rock images. Results show that a 3-layer network obtains an average accuracy of 99.60% across 10 trials on the test set. A version of Self-taught Learning was also implemented to prove the generalizability of the features extracted by the CNN. Finally, one model has been chosen to be deployed on a mobile device to demonstrate practicality and portability. The deployed model achieves a perfect classification accuracy on the test set, while taking only 0.068 seconds to make a prediction, equivalent to about 14 frames per second.

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.

Radiator reliability is crucial in environments characterized by high temperatures and friction, where prompt interventions are often required to prevent system failures. This study introduces a proactive approach to radiator fault diagnosis, leveraging the integration of Gaussian Mixture Model (GMM) and Long-Short Term Memory (LSTM) autoencoders. Vibration signals from radiators were systematically collected through randomized durability vibration bench tests, resulting in four operating states—two normal, one unknown, and one faulty. Time-domain statistical features of these signals were extracted and subjected to Principal Component Analysis (PCA) to facilitate efficient data interpretation. Subsequently, this study discusses the comparative effectiveness of GMM and LSTM in fault detection. GMMs are deployed for initial fault classification, leveraging their clustering capabilities, while LSTM autoencoders excel in capturing time-dependent sequences, facilitating advanced anomaly detection for previously unencountered faults. This alignment offers a potent and adaptable solution for radiator fault diagnosis, particularly in challenging high-temperature or high-friction environments. Consequently, the proposed methodology not only provides a robust framework for early-stage fault diagnosis but also effectively balances diagnostic capabilities during operation. Additionally, this study presents the foundation for advancing reliability life assessment in accelerated life testing, achieved through dynamic threshold adjustments using both the absolute log-likelihood distribution of the GMM and the reconstruction error distribution of the LSTM autoencoder model.

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.

Neurodegenerative illnesses, such as Parkinson’s disease (PD), have a substantial impact on the overall well-being of those who are affected. This study investigates and contrasts the capabilities of convolutional neural networks (CNN) in detecting Parkinson’s disease (PD) by utilising hand-drawn images alongside wave and spiral images as input data. This study employs pre-trained CNN models, specifically MobileNet, ResNet50, EfficientNet-B1, and InceptionV3, to classify Parkinson’s disease (PD). The findings demonstrate that MobileNet surpasses other architectural designs, as evidenced by the F1-Score of the four classes: Spiral Normal (0.87), Spiral Parkinson (0.86), Wave Normal (0.97), and Wave Parkinson (0.97). MobileNet has also shown a remarkable accuracy of 0.92 in diagnosing Parkinson’s disease. The result demonstrates the efficacy of MobileNet in extracting features from images. The results of this study enhance the application of deep learning methods in the early detection of PD, as well as help indicate the effectiveness of patient therapy and exercise, promising better patient outcomes through timely intervention and treatment.

Estimating plume cloud height is essential for various applications, such as global climate models. Smokestack plume rise is the constant height at which the plume cloud is carried downwind as its momentum dissipates and the plume cloud and the ambient temperatures equalize. Although different parameterizations are used in most air-quality models to predict the plume rise, they have been unable to estimate it properly. This paper proposes a novel framework to monitor smokestack plume clouds and make long-term, real-time measurements of the plume rise. For this purpose, a three-stage framework is developed based on Deep Convolutional Neural Networks (DCNNs). In the first stage, an improved Mask R-CNN, called Deep Plume Rise Network (DPRNet), is applied to recognize the plume cloud. Then, image processing analysis and least squares theory are respectively used to detect the plume cloud’s boundaries and fit an asymptotic model into their centerlines. The y-component coordinate of this model’s critical point is considered the plume rise. In the last stage, a geometric transformation phase converts image measurements into real-life ones. A wide range of images with different atmospheric conditions, including day, night, and cloudy/foggy, have been selected for the DPRNet training algorithm. Obtained results show that the proposed method outperforms widely-used networks in smoke border detection and recognition.

Wideband signal detection is an important problem in wireless communication. With the rapid development of deep learning (DL) technology, some DL-based methods are applied to wireless technology, and the effect is obvious. In this paper, we propose a novel neural network for multi-type signal detection that can locate signals and recognize signal types in wideband spectrogram. Our network utilizes the key point estimation to locate the rough centerline of signal region and identify class. Then, several regressions are carried out to achieve properties, such as local offset and border offsets of bounding box, which is further synthesized for a more fine location. Experimental results demonstrate that our method performs more accurate than other DL-based object detection methods previously employed for the same detection task. Specifically, our method runs obviously faster than existing methods, and abandons the anchor generation, which makes it more favorable for real-time applications.

Lamb wave-based structural health monitoring is widely acknowledged as a reliable 11 method for damage identification, classification, localization and quantification. However, due to 12 the complexity of Lamb wave signals, especially after interacting with structural components and 13 defects, interpreting these waves and extracting useful information about the structure's health is 14 still a major challenge. However, deep learning-based strategy offers a great opportunity to address 15 such challenges as the algorithm can operate directly on raw discrete time-domain signals. Unlike 16 traditional methods, which often require careful feature engineering and preprocessing, deep learn-17 ing can automatically extract relevant features from the raw data. This paper proposes an autoen-18 coder based on a bidirectional long short-term memory network with maximal overlap discrete 19 wavelet transform layer to detect the signal anomaly and determine the location of the damage in 20 the composite structure. This approach has the potential to greatly enhance our ability to detect and 21 locate structural damage in composite structures, thereby increasing safety and efficiency.

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.

To solve two key problems in the identification of people who are infected with COVID-19: the first problem is that the identification accuracy is not high enough. The second problem is that present identification method such as nucleic acid testing is expensive in many countries. Methods: So, I decided to design a fast identification method for COVID-19 patients which is based on deep learning. After the model (CoughNet) learns more than 6,000 cough spectrograms of both COVID-19 patients and normal people, the accuracy rate of identification of COVID-19 patients and normal people is higher than 99% in the test set. Structure: This paper is mainly divided into three parts: the first part introduces the main background and research status of the research; The second part introduces the research methods; The third part introduces the specific process of the experiment.

All kinds of vessels consist of dozens of complex machineries with rotating parts and electric motors that operate continuously in a harsh environment with excess temperature and humidity, vibration, fatigue and load. A breakdown or malfunction in one of these machineries can significantly impact the vessel’s operation and safety and consequently, the safety of the crew and the environment. To maintain operational efficiency and seaworthiness, the shipping industry invests substantial resources in preventive maintenance and repairs. This research presents the economic and technical benefits of predictive maintenance over to traditional preventive maintenance, and repair by replacement approaches in the maritime domain. By leveraging modern technology and Artificial Intelligence, we can analyze real-time operating conditions of machinery, enabling early detection of potential damages and allowing for effective planning of future maintenance and repair activities. In this paper, we propose and develop a Convolutional Neural Network that is fed with raw vibration measurements acquired in a laboratory environment from the ball bearings of a motor. Then, we investigate whether the proposed network can accurately detect the functional state of ball bearings and categorize any possible failures present, contributing to improved maintenance practices in the shipping industry.

Modern energy systems collect high volumes of data that can provide valuable information about energy consumption. Electric companies can now use historical data to make informed decisions on energy production by forecasting the expected demand. Many deep learning models have been proposed to deal with these type of time series forecasting problems. Deep neural networks, such as recurrent or convolutional, can automatically capture complex patterns in time series data and provide accurate predictions. In particular, Temporal Convolutional Networks (TCN) are a specialised architecture that has advantages over recurrent networks for forecasting tasks. TCNs are able to extract long-term patterns using dilated causal convolutions and residual blocks, and can also be more efficient in terms of computation time. In this work, we propose a TCN-based deep learning model to improve the predictive performance in energy demand forecasting. Two energy-related time series with data from Spain have been studied: the national electric demand, and the power demand at charging stations for electric vehicles. An extensive experimental study has been conducted, involving more than 1900 models with different architectures and parametrisations. The TCN proposal outperforms the forecasting accuracy of Long Short-Term Memory (LSTM) recurrent networks, which are considered the state-of-the-art in the field.

Anomaly occurrences in hydraulic machinery may lead to massive systems shut down, jeopardizing the safety of the machinery and its surrounding human operator(s) and environment, and the severe economic implications succeeding the faults and their associated damage. Hydraulics are mostly placed in ruthless environments, where they are consistently vulnerable to many faults. Hence, not only the machines and their components are prone to anomalies, but also the sensors attached to them, which monitor and report their health and behavioral changes. In this work, a comprehensive applicational analysis of anomalies in hydraulic systems extracted from a hydraulic test rig is thoroughly achieved. Firstly, we provided a combination of a new architecture of LSTM autoencoders and supervised machine and deep learning methodologies to perform two separate stages of fault detection and diagnosis. The two phases are condensed by: the detection phase using the LSTM autoencoder. Followed by the fault diagnosis phase represented by the classification schema. The previously mentioned framework is applied to both component and sensor faults in hydraulic systems, deployed in the form of two in-depth applicational experiments. Moreover, a thorough literature review of the past decade related work for the two stages separately is successfully conducted in this paper.

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

Accurately mapping individual tree species in densely forested environments is crucial to forest inventory. When considering only RGB images, this is a challenging task for many automatic photogrammetry processes. The main reason for that is the spectral similarity between species in RGB scenes, which can be a hindrance for most automatic methods. State-of-the-art deep learning methods could be capable of identifying tree species with an attractive cost, accuracy, and computational load in RGB images. This paper presents a deep learning-based approach to detect an important multi-use species of palm trees (Mauritia flexuosa; i.e., Buriti) on aerial RGB imagery. In South-America, this palm tree is essential for many indigenous and local communities because of its characteristics. The species is also a valuable indicator of water resources, which comes as a benefit for mapping its location. The method is based on a Convolutional Neural Network (CNN) to identify and geolocate singular tree species in a high-complexity forest environment, and considers the likelihood of every pixel in the image to be recognized as a possible tree by implementing a confidence map feature extraction. This study compares the performance of the proposed method against state-of-the-art object detection networks. For this, images from a dataset composed of 1,394 airborne scenes, where 5,334 palm-trees were manually labeled, were used. The results returned a mean absolute error (MAE) of 0.75 trees and an F1-measure of 86.9%. These results are better than both Faster R-CNN and RetinaNet considering equal experiment conditions. The proposed network provided fast solutions to detect the palm trees, with a delivered image detection of 0.073 seconds and a standard deviation of 0.002 using the GPU. In conclusion, the method presented is efficient to deal with a high-density forest scenario and can accurately map the location of single species like the M flexuosa palm tree and may be useful for future frameworks.

In recent years, Edge Intelligence (EI) has emerged, combining edge computing with AI, specifically deep learning, to run AI algorithms directly on edge devices. In practical applications, EI faces challenges related to computational power, power consumption, size, and cost, with the primary challenge being the trade-off between computational power and power consumption. This has rendered traditional computing platforms unsustainable, making heterogeneous parallel computing platforms a crucial pathway for implementing EI. In our research, we leveraged the Xilinx Zynq 7000 heterogeneous computing platform, employed High-Level Synthesis (HLS) for design, and implemented two different accelerators for LeNet-5 using loop unrolling and pipelining optimization techniques. The experimental results show that when running at a clock speed of 100 MHz, the PIPELINE accelerator, compared to the UNROLL accelerator, experiences an 8.09% increase in power consumption but speeds up by 14.972 times, making the PIPELINE accelerator superior in performance. Compared to the CPU, the PIPELINE accelerator reduces power consumption by 91.37% and speeds up by 70.387 times, while compared to the GPU, it reduces power consumption by 93.35%.This study provides two different optimization schemes for edge intelligence applications through design and experimentation, and demonstrates the impact of different quantization methods on FPGA resource consumption. These experimental results can provide a reference for practical applications, thereby providing a reference hardware acceleration scheme for edge intelligence applications.

An application of deep convolutional neural network and recurrence plot for financial market movement prediction is presented. Though it is challenging and subjective to interpret its information, the pattern formed by a recurrence plot provide a useful insight into the dy- namical system. We used a recurrence plot of seven financial time series to train a deep neural network for financial market movement predic- tion. Our approach is tested on our dataset and achieved an average of 53.25% classification accuracy. The result suggests that a well trained deep convolutional neural network can learn a recurrence plot and pre- dict a financial market direction.

3D object detection is crucial for autonomous driving to understand the driving environment. Since the pooling operation causes information loss in the standard CNN, we have designed a wavelet multiresolution analysis-based 3D object detection network without a pooling operation. Additionally, instead of using a single filter like the standard convolution, we use the lower-frequency and higher-frequency coefficients as a filter. These filters capture more relevant parts than a single filter, enlarging the receptive field. The model comprises a discrete wavelet transform (DWT) and an inverse wavelet transform (IWT) with skip connections to encourage feature reuse for contrasting and expanding layers. The IWT enriches the feature representation by fully recovering the lost details during the downsampling operation. Element-wise summation is used for the skip connections to decrease the computational burden. We train the model for the Haar and Daubechies (Db4) wavelets. The two-level wavelet decomposition result shows that we can build a lightweight model without losing significant performance. The experimental results on the KITTI’s BEV and 3D evaluation benchmark show our model outperforms the Pointpillars base model by up to 14 \% while reducing the number of trainable parameters. Code will be released.

State-of-the-art remote sensing scene classification methods employ different Convolutional Neural Network architectures for achieving very high classification performance. A trait shared by the majority of these methods is that the class associated with each example is ascertained by examining the activations of the last fully connected layer, and the networks are trained to minimize the cross-entropy between predictions extracted from this layer and ground-truth annotations. In this work, we extend this paradigm by introducing an additional output branch which maps the inputs to low dimensional representations, effectively extracting additional feature representations of the inputs. The proposed model imposes additional distance constrains on these representations with respect to identified class representatives, in addition to the traditional categorical cross-entropy between predictions and ground-truth. By extending the typical cross-entropy loss function with a distance learning function, our proposed approach achieves significant gains across a wide set of benchmark datasets in terms of classification, while providing additional evidence related to class membership and classification confidence.

In the field of maneuvering target tracking, the combined observations of azimuth and Doppler may cause weak observation or non-observation in the application of traditional target tracking algorithms. Additionally, traditional target-tracking algorithms require pre-defined multiple mathematical models to accurately capture the complex motion states of targets, while model mismatch and unavoidable measurement noise lead to significant errors in target state prediction. To address those above challenges, in recent years, the target-tracking algorithms based on neural networks, such as recurrent neural networks (RNNs), long short-term memory (LSTM) networks, and Transformer architectures, have been widely used for their unique advantages to achieve accurate predictions. To better model the nonlinear relationship between the observation time series and the target state time series, as well as the contextual relationship among time series points, we present a deep learning algorithm called recursive downsample-convolve-interact neural network (RDCINN) based on convolutional neural network (CNN) that downsamples time series into sub-sequences and extracts multi-resolution features to enable the modeling of complex relationships between time series, which overcomes the shortcomings of traditional target-tracking algorithms in using observation information inefficiently due to weak observation or non-observation. The experimental results show that our algorithm outperforms other existing algorithms in the scenario of strong maneuvering target tracking with the combined observations of azimuth and Doppler.

Rice has become the second major food among rural people. It is one of the most nutrient and low-cost food available in Asia. Rice is composed of two major subspecies as Japonica and Indica and it belongs to the Poaceae family. The World Bank predicts that rice demand will grow 51% faster than population growth in 2025. It is high time to develop a model which predicts the type of infected paddy leaf and avoids the loss of grains. In this paper, the proposed hybrid CNN model, three CNN networks with different layer properties are trained parallel. The extracted feature from these CNN networks is combined to provide the classification of the type of infected leaf. Fieldwork was carried and about 340 samples are collected from five different types of classes. The proposed hybrid model was tested with both the fieldwork dataset and also with an available dataset on the internet. The experimental results show that the proposed model outperforms the other CNN network and yields an accuracy value of 98%, a precision value of 95%, and a misclassification rate of 4% respectively.

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.

Tail-biting convolutional codes extend the classical zero-termination convolutional codes: Both encoding schemes force the equality of start and end states, but under the tail-biting each state is a valid termination. This paper proposes a machine-learning approach to improve the state-of-the-art decoding of tail-biting codes, focusing on the widely employed short length regime as in the LTE standard. This standard also includes a CRC code. First, we parameterize the circular Viterbi algorithm, a baseline decoder that exploits the circular nature of the underlying trellis. An ensemble combines multiple such weighted decoders, each decoder specializes in decoding words from a specific region of the channel words' distribution. A region corresponds to a subset of termination states; the ensemble covers the entire states space. A non-learnable gating satisfies two goals: it filters easily decoded words and mitigates the overhead of executing multiple weighted decoders. The CRC criterion is employed to choose only a subset of experts for decoding purpose. Our method achieves FER improvement of up to 0.75dB over the CVA in the waterfall region for multiple code lengths, adding negligible computational complexity compared to the circular Viterbi algorithm in high SNRs.

In recent years, there has been a growing interest in the prediction of facial age due to its diverse applications in fields such as security, entertainment, and healthcare. The current scourge of underaged voting in Nigeria is a problem and this research delves into the realm of facial age prediction by employing four well-known convolutional neural network (CNN) architectures, namely VGG16, ResNet50, Mobile Net, and VGG19 to predict chronological ages from facial pictures of person. The objective is to achieve precise age estimation from facial images, utilizing two datasets: UTKFace and CASIA African Facial Datasets. The results of this investigation are noteworthy. Specifically, the VGG16 model which demonstrated remarkable performance, yielding a Mean Absolute Error (MAE) of 1.76 when applied to the UTK-Face dataset. Additionally, when utilizing Mobile-Net, an unprecedented MAE of 1.10 was achieved for the Casia Africa Face Dataset. Notably, this marks the first instance of employing the dataset for facial age detection with CNNs, and this approach outperformed previous works, yielding the lowest MAE among all the studies reviewed.

Generative AI has gained enormous interest nowadays due to new applications like chatGPT, DALL E, Stable Difussion and Deep Fake. Particularly DALL E, Stable Difussion and others (Adobe Firefly, ImagineArt...) are able to create images from a text prompt and are also able to recreate real photographs. Due to this fact, intense research has arisen to create new image forensics applications able to distinguish between real captured images and videos and artificial ones. Detecting forgeries made with Deep Fake is one of the most researched issues. This paper is about another kind of forgery detection. The purpose of this research aims to detect photo realistic AI created images versus real photos coming from a physical camera. For this purpose, techniques that perform a pixel level feature extraction are used. First one is Photo Response Non-Uniformity (PRNU). PRNU is a special noise due to imperfections on the camera sensor that is used for source camera identification. The underlying idea is that AI images will have a different PRNU pattern. Second one is Error level analysis (ELA). This is other type of feature extraction traditionally used for detecting image editions. In fact, ELA is being used nowadays by photographers to detect manually AI created images. Both kinds of features are used to train Convolutional Neural Networks to differentiate between AI images and real photographs. Good results are obtained achieving accuracy rates over 95%. Both extraction methods are carefully assessed by computing precision/recall and F1-score measurements.

Clients are increasingly looking for fast and effective means to quickly and frequently survey and communicate the condition of their buildings so that essential repairs and maintenance work can be done in a proactive and timely manner before it becomes too dangerous and expensive. Traditional methods for this type of work commonly comprise of engaging building surveyors to undertake a condition assessment which involves a lengthy site inspection to produce a systematic recording of the physical condition of the building elements, including cost estimates of immediate and projected long-term costs of renewal, repair and maintenance of the building. Current asset condition assessment procedures are extensively time consuming, laborious, and expensive and pose health and safety threats to surveyors, particularly at height and roof levels which are difficult to access. We propose a method for automated detection and localisation of key building defects from images using deep learning and convolution neural networks. The proposed model is based on a pre-trained VGG-16 classifier with Class Activation Mapping (CAM) for object localisation. The model has proven to be robust and able to accurately detect and localise mould growth, stains, and paint deterioration defects arising from dampness in buildings. The approach is being developed with potentials to scale-up to support automated detection of defects and deterioration of buildings in real-time using mobile devices and drones.

Automatic garments design class identification for recommending the fashion trends is important nowadays because of the rapid growth of online shopping. By learning the properties of images efficiently, a machine can give better accuracy of classification. Several methods, based on Hand-Engineered feature coding exist for identifying garments design classes. But, most of the time, those methods do not help to achieve better results. Recently, Deep Convolutional Neural Networks (CNNs) have shown better performances for different object recognition. Deep CNN uses multiple levels of representation and abstraction that helps a machine to understand the types of data (images, sound, and text) more accurately. In this paper, we have applied deep CNN for identifying garments design classes. To evaluate the performances, we used two well-known CNN models AlexNet and VGGNet on two different datasets. We also propose a new CNN model based on AlexNet and found better results than existing state-of-the-art by a significant margin.

The detection of seismic events at regional and teleseismic distances is critical to Nuclear Treaty Monitoring. Traditionally, detecting regional and teleseismic events has required the use of an expensive multi-instrument seismic array; however in this work, we present DeepPick, a novel seismic detection algorithm capable of array-like performance from a single trace. We achieve this directly, by training our single-trace detector against labeled events from an array catalog, and by utilizing a deep temporal convolutional neural network. The training data consists of all arrivals in the International Seismological Centre Catalog for seven seismic arrays over a five year window from 1 Jan 2010 to 1 Jan 2015, yielding a total training set of 608,362 detections. The test set consists of the same seven arrays over a one year window from 1 Jan 2015 to 1 Jan 2016. We report our results by training the algorithm on six of the arrays and testing it on the seventh, so as to demonstrate the transportability and generalization of the technique to new stations. Detection performance against this test set is outstanding. Fixing a type-I error rate of 1%, the algorithm achieves an overall recall rate of 73% on the 141,095 array beam picks in the test set, yielding 102,394 correct detections. This is more than 4 times the 23,259 detections found in the analyst-reviewed single-trace catalogs over the same period, and represents an 8dB improvement in detector sensitivity over current methods. These results demonstrate the potential of our algorithm to significantly enhance the effectiveness of the global treaty monitoring network.

Social networks have captured the attention of many people worldwide. However, these services have also attracted a considerable number of malicious users whose purpose is to compromise digital assets of other members by using messages as an attack vector to execute different variants of cyberattacks against them. Therefore, this work presents an approach based on Natural Language Processing tools and a Convolutional Neural Network architecture to detect and classify, on social network messages, four types of cyberattacks, such as malware, phishing, spam, and even one whose purpose is deceiving the user into spreading malicious messages to other users, which in this work is identified as bot attacks. One notable feature of this work is that it analyzes textual content without depending on any characteristics from a specific social network, making its analysis independent from particular data sources. Finally, this work was tested on real data, demonstrating its results in two stages. The first detects the existence of any of the four cyberattacks within the message, obtaining an accuracy value of 0.91. After detecting a message as a cyberattack, the next stage is to classify it into one of the four types of cyberattack, achieving an accuracy value of 0.82.

With the widespread of embedded sensing capabilities of mobile devices, there has been unprecedented development of context-aware solutions. This allows the proliferation of various intelligent applications such as those for remote health and lifestyle monitoring, intelligent personalized services, etc. However, activity context recognition based on multivariate time series signals obtained from mobile devices in unconstrained conditions is naturally prone to imbalance class problems. This means that recognition models tend to predict classes with the majority number of samples whilst ignoring classes with the least number of samples, resulting in poor generalization. To address this problem, we propose to augment the time series signals from inertia sensors with signals from ambient sensing to train deep convolutional neural networks (DCNN) models. DCNN provides the characteristics that capture local dependency and scale invariance of these combined sensor signals. Consequently, we developed a DCNN model using only inertial sensor signals and then developed another model that combined signals from both inertia and ambient sensors aiming to investigate the class imbalance problem by improving the performance of the recognition model. Evaluation and analysis of the proposed system using data with imbalanced classes show that the system achieved better recognition accuracy when data from inertial sensors are combined with those from ambient sensors such as environment noise level and illumination, with an overall improvement of 5.3% accuracy.

Deepfakes are manipulated or altered images, or video, that are created using deep learning models with high levels of photorealism. The two popular methods of producing a deepfake are based on either convolutional neural networks (CNN), or autoencoders. Deepfakes created using CNN comparatively show higher qualities of realism, yet oftentimes leave artifacts and distortions in the generated media that can be detected using machine learning and deep learning algorithms. In recent years, there has been an influx of periocular image and video data because of the increase usage of face masks. By wearing masks, much of what is used for facial recognition is hidden, leaving only the periocular region visible to an observer. This loss of vital information leads to easier misidentification of media, allowing deepfakes to less likely be identified as fake. In this work, feature extraction methods, such as Scale-Invariant Feature Transform (SIFT), Histogram of Oriented Gradients (HOG), and CNN, are used to train an ensemble deep learning model to detect deepfakes in videos on a frame-by-frame level based on the periocular region. Our proposed model is able to distinguish original and manipulated images with accuracies around 98.9 percent, which is an improvement to previous works by combining SIFT and HOG for deepfake detection in convolutional neural networks.

Polar mesocyclones (MCs) are small marine atmospheric vortices. The class of intense MCs, called polar lows, are accompanied by extremely strong surface winds and heat fluxes and thus largely influencing deep ocean water formation in the polar regions. Accurate detection of polar mesocyclones in high-resolution satellite data, while challenging, is a time-consuming task, when performed manually. Existing algorithms for the automatic detection of polar mesocyclones are based on the conventional analysis of patterns of cloudiness and involve different empirically defined thresholds of geophysical variables. As a result, various detection methods typically reveal very different results when applied to a single dataset. We develop a conceptually novel approach for the detection of MCs based on the use of deep convolutional neural networks (DCNNs). As a first step, we demonstrate that DCNN model is capable of performing binary classification of 500x500km patches of satellite images regarding MC patterns presence in it. The training dataset is based on the reference database of MCs manually tracked in the Southern Hemisphere from satellite mosaics. We use a subset of this database with MC diameters falling in the range of 200-400 km. This dataset is further used for testing several different DCNN setups, specifically, DCNN built “from scratch”, DCNN based on VGG16 pre-trained weights also engaging the Transfer Learning technique, and DCNN based on VGG16 with Fine Tuning technique. Each of these networks is further applied to both infrared (IR) and a combination of infrared and water vapor (IR+WV) satellite imagery. The best skills (97% in terms of the binary classification accuracy score) is achieved with the model that averages the estimates of the ensemble of different DCNNs. The algorithm can be further extended to the automatic identification and tracking numerical scheme and applied to other atmospheric phenomena characterized by a distinct signature in satellite imagery.

Preliminaries Convolutional Neural Network (CNN) applications have recently emerged in Structural Health Monitoring (SHM) systems focusing mostly on vibration analysis. However, the SHM literature shows clearly that there is a lack of application regarding the combination of PZT (Lead Zirconate Titanate) based method and CNN. Likewise, applications using CNN along with the Electromechanical Impedance (EMI) technique applied to SHM systems are rare. To encourage this combination, an innovative SHM solution through the combination of the EMI-PZT and CNN is presented here. To accomplish this, the EMI signature is split into several parts followed by computing the Euclidean distances among them to form a RGB (red, green and blue) frame. As a result, we introduce a dataset formed from the EMI-PZT signals of 720 frames, encompassing a total of 4 types of structural conditions for each PZT. In a case study, the CNN-based method was experimentally evaluated using three PZTs glued onto an aluminum plate. The results reveal an effective pattern classification; yielding a 100% hit rate which outperforms other SHM approaches. Furthermore, the method needs only a small dataset for training the CNN, providing several advantages for industrial applications.

In order to design effective protective coatings against corrosion, the polyvinyl alcohol (PVA) as compound and composite with silver nanoparticles (nAg/PVA) were electrodeposited on copper surface employing electrochemical techniques such as linear potentiometry and cyclic voltammetry. A new paradigm was used to distinguish the features of coatings, i.e., a Deep Convolutional Neural Network (CNN) was implemented to automatically and hierarchically extract the discriminative characteristics from the information given by optical microscopy images. The main arguments that invoke a CNN implementation in the surface science of materials are the following: artificial intelligence techniques can be successfully applied to learn differences between surface coatings; based on their popularity for image processing, CNN can model images related to the problem of coatings; deep learning is able to extract the features that are distinguishable between material surfaces. To provide an overview of the copper surface, CNN was applied on microscope slides (CNN@microscopy) and inherently learnt distinctive characteristics for each class of surface morphology. The material surface morphology controlled by CNN without the interference of the human factor was successfully conducted, in our study, to extract the similarities/differences between unprotected and protected surfaces to establish the PVA and nAg/PVA performance to retard the copper corrosion.

This paper presents a novel approach for semantic segmentation of building roof in dense urban environment with Deep Convolution Neural Network (DCNN) using imagery acquired by a Chinese Very High Resolution (VHR) satellite mission, i.e. GaoFen-2 (GF-2). To provide an operational end-to-end work flow for accurate build roof mapping with feature extraction as well as image segmentation, a fully convolutional DCNN with both convolutional and deconvolutional layers is designed to perform the VHR image analysis for labeling pixels. Since the diverse urban patterns and building styles in large areas, sample image data sets of building roof and non-building roof are collected over different metropolitan regions in China. We selected typical cities with dense urban environment in each metropolitan region as study areas for collecting training and test samples. High performance cluster with GPU-mounted workstations is employed to perform the model training and optimization. With the building roof samples collected over different cities, the predictive model with multiple NN layers is developed for building roof labeling. The validation of the building roof map shows that the overall accuracy(OA) and the mean Intersection Over Union( mIOU) of DCNN based segmentation are 94.67%, 0.85 respectively, while CRF-refined segmentation achieved OA of 94.69% and mIOU of 0.83. The results suggest that the proposed approach is a promising solution for building roof mapping with VHR images over large areas across different urban and building patterns. With the operational acquisition of GF2 VHR imagery, it is expected to develop an automated pipeline for operational built-up area monitoring and timely update of building roof map over large areas.

As the world increasingly becomes more interconnected, the demand for safety and security is ever-increasing, particularly for industrial networks. This has prompted numerous researchers to investigate different methodologies and techniques suitable for intrusion detection systems (IDS) requirements. Over the years, many studies have proposed various solutions in this regard including signature-based and machine-learning (ML) based systems. More recently, researchers are considering deep learning (DL) based anomaly detection approaches. Most proposed works in this research field aimed to achieve either one or a combination of high accuracy, considerably low false alarm rates (FARs), high classification specificity and detection sensitivity, achieving lightweight DL models, or other ML and DL-related performance measurement metrics. In this study, we propose a novel method to convert a raw dataset to an image dataset to magnify patterns. Based on this we devise an anomaly detection for IDS using a lightweight convolutional neural network (CNN) that classifies denial of service and distributed denial of service. The proposed methods were evaluated using a modern dataset, CSE-CIC-IDS2018, and a legacy dataset, NSL-KDD. We have also applied a combined dataset to assess the generalization of the proposed model across various datasets. Our experimental results have demonstrated that the proposed methods achieved high accuracy and considerably low FARs with high specificity and sensitivity. The resulting loss and accuracy curves have also demonstrated the excellent generalization of the proposed lightweight CNN model, effectively avoiding overfitting. This holds for both the modern and legacy datasets, including their mixed version.

Substances that can modify the androgen receptor pathway in humans and animals are entering the environment and food chain with the proven ability to disrupt hormonal systems and leading to toxicity and adverse effects on reproduction, brain development, and prostate cancer, among others. State-of-the-art databases with experimental data of human, chimp, and rat effects by chemicals have been used to build machine learning classifiers and regressors and evaluate these on independent sets. Different featurizations, algorithms, and protein structures lead to dif- ferent results, with deep neural networks (DNNs) on user-defined physicochemically-relevant features developed for this work outperforming graph convolutional, random forest, and large featurizations. The results show that these user-provided structure-, ligand-, and statistically-based features and specific DNNs provided the best results as determined by AUC (0.87), MCC (0.47), and other metrics and by their interpretability and chemical meaning of the descriptors/features. In addition, the same features in the DNN method performed better than in a multivariate logistic model: validation MCC = 0.468 and training MCC = 0.868 for the present work compared to evalu- ation set MCC = 0.2036 and training set MCC = 0.5364 for the multivariate logistic regression on the full, unbalanced set. Techniques of this type may improve AR and toxicity description and predic- tion, improving assessment and design of compounds. Source code and data are available at https://github.com/AlfonsoTGarcia-Sosa/ML

Accurate and efficient traffic incident detection methods can effectively alleviate traffic congestion caused by traffic incidents, prevent secondary accidents and improve the safety of urban road traffic.Aiming at the problems that the traditional machine learning event detection method cannot fully extract the parameter characteristics of traffic flow and is not suitable formulti-dimensional and non-linear massive data, we propose a new traffic event detection method(CNN-XGBoost).This method combines the respective advantages of Convolution Neural Network(CNN) and Extreme Gradation Boosting (XGBoost). Firstly, we preprocessed the original freeway traffic incident detection data set by constructing initial variable set, data normalization, data balance processing and dimension reorganization. Secondly,we use CNN network to automatically extract the deep features of event detection data, and use XGBoost as a classifier to classify the extracted features for expressway traffic event detection.Finally, we use the data set of Hangzhou expressway microwave detector in China to carry out simulation experiments on CNN-XGBoost. The experimental results show that compared with XGBoost, CNN, Support Vector Machine (SVM) and Gradient Boosting Decision Tree (GBDT) and other methods, CNN-XGBoost method can effectively improve the accuracy of expressway traffic event detection and has better generalization ability.

It is a challenging problem that air-written Arabic letters has received a lot of attention in the past decades when compared to commonly spoken languages like English languages. To fill this gap, we propose a strong model that brings together machine learning (ML) and optical character recognition (OCR) methods. The model applied several ML methods, (i.e., Neural Networks (NN), Random Forest (RF), K-Nearest Neighbors (KNN), and Support Vector Machine (SVM), with deep convolutional neural networks (CNNs) such as VGG16, VGG19, and SqueezeNet for effective feature extraction. Our study utilizes the AHAWP dataset, which consists of varied writing styles and variations in hand signs, to train and evaluate the model. Preprocessing systems are applied to improve data quality by reduction bias. Besides, OCR methods are combined into our model to sequestrate individual letters from continuous air-written gestures and refine recognition results. Results of this study show that the proposed model has achieved the extreme accuracy of 88.8% using NN with VGG16. This study presents an inclusive approach that combines ML, deep CNNs, and OCR methods to address the issue of Arabic in air writing recognition research.

Deep learning has shown remarkable results in every field, especially in the biomedical field, due to its ability to exploit large-scale datasets. A convolutional neural network (CNN) is a widely used deep learning approach to solve medical imaging problems. Over the past few years, many studies have focused on CNN-based techniques for brain tumor diagnosis. There are, however, still some critical challenges that CNNs face towards clinic application. This study presents a comprehensive review of current literature that involves CNN architectures for brain tumor classification. We compare the key achievements in the performance evaluation metrics of the applied classification algorithms. In addition, this review assesses the clinical effectiveness of the included studies to elaborate on the limitations and directions of this area for future work. No review focusing on the clinical effectiveness of previous works in this field has been published. We believe that this study has the potential to elevate the application of CNN-based deep learning methods in clinical practice and also can be a quick reference for biomedical researchers who are interested in this field.

Automated lip reading is a research problem that has developed considerably in recent years. Lip reading is evaluated both visually and audibly in some cases. The lip reading model is a field of use for detecting specific words using images from security cameras, but it is not possible to use audio-visual databases in this situation. It is not possible to obtain the sound input of the pronounced word in all cases. We collected a new Turkish dataset with only the image in this study. The new dataset is produced using Youtube videos, which is an uncontrolled environment. For this reason, images have difficult parameters in terms of environmental factors such as light, angle, color, and personal characteristics of the face. Despite the different features on the human face such as mustache, beard, and make-up, the visual speech recognition problem was developed on 10 classes including single words and two-word phrases using Convolutional Neural Networks (CNN) without any intervention on the data. The proposed study using only-visual data obtained a model which is automated visual speech recognition with a deep learning approach. In addition, since this study uses only-visual data, the computational cost and resource usage is less than in multi-modal studies. It is also the first known study to address the lip reading problem with a deep learning algorithm using a new dataset belonging to the Ural-Altaic languages.

As a relatively high percentage of adenoma polyps are missed, a computer-aided diagnosis (CAD) tool based on deep learning can aid the endoscopist in diagnosing colorectal polyps or colorectal cancer in order to decrease polyps missing rate and prevent colorectal cancer mortality. Convolutional Neural Network (CNN) is a deep learning method and has achieved better results in detecting and segmenting specific objects in images in the last decade than conventional models such as regression, support vector machines or artificial neural networks. In recent years, based on the studies in medical imaging criteria, CNN models have acquired promising results in detecting masses and lesions in various body organs, including colorectal polyps. In this review, the structure and architecture of CNN models and how colonoscopy images are processed as input and converted to the output are explained in detail. In most primary studies conducted in the colorectal polyp detection and classification field, the CNN model has been regarded as a black box since the calculations performed at different layers in the model training process have not been clarified precisely. Furthermore, I discuss the differences between the CNN and conventional models, inspect how to train the CNN model for diagnosing colorectal polyps or cancer, and evaluate model performance after the training process.

Although automated Acute Lymphoblastic Leukemia (ALL) detection is essential, it is challenging due to the morphological correlation between malignant and normal cells. The traditional ALL classification strategy is arduous, time-consuming, often suffers inter-observer variations, and necessitates experienced pathologists. This article has automated the ALL detection task, employing deep Convolutional Neural Networks (CNNs). We explore the weighted ensemble of deep CNNs to recommend a better ALL cell classifier. The weights are estimated from ensemble candidates' corresponding metrics, such as accuracy, F1-score, AUC, and kappa values. Various data augmentations and pre-processing are incorporated for achieving a better generalization of the network. We train and evaluate the proposed model utilizing the publicly available C-NMC-2019 ALL dataset. Our proposed weighted ensemble model has outputted a weighted F1-score of 88.6%, a balanced accuracy of 86.2%, and an AUC of 0.941 in the preliminary test set. The qualitative results displaying the gradient class activation maps confirm that the introduced model has a concentrated learned region. In contrast, the ensemble candidate models, such as Xception, VGG-16, DenseNet-121, MobileNet, and InceptionResNet-V2, separately produce coarse and scatter learned areas for most example cases. Since the proposed ensemble yields a better result for the aimed task, it can experiment in other domains of medical diagnostic applications.

In recent years, the demand for coffee has increased tremendously. During the production process, green coffee beans are traditionally screened manually for defective beans before they are packed into coffee bean packages; however, this method is not only time-consuming but also increases the rate of human error due to fatigue. Therefore, this paper proposed a lightweight deep convolutional neural network (LDCNN) for the quality detection system of green coffee beans, which combined depthwise separable convolution (DSC), squeeze-and-excite block (SE block), skip block, and other frameworks. To avoid the influence of low parameters of the lightweight model caused by the model training process, rectified Adam (RA), lookahead (LA), and gradient centralization (GC) were included to improve efficiency; the model was also put into the embedded system. Finally, the local interpretable model-agnostic explanations (LIME) model was employed to explain the predictions of the model. The experimental results indicated that the accuracy rate of the model could reach up to 98.38% and the F1 score could be as high as 98.24% when detecting the quality of green coffee beans. Hence, it can obtain higher accuracy, lower computing time, and lower parameters. Moreover, the interpretable model verified that the lightweight model in this work is reliable, providing the basis for screening personnel to understand the judgment through its interpretability, thereby improving the classification and prediction of the model.

The administration of regional neuraxial analgesia for pain relief during labor is widely recognized as a safe and effective method involving medication delivery into the epidural or subarachnoid space in the lower back. This study proposes an innovative semantic image segmentation methodology emphasizing enhanced interpretability using convolutional Random Fourier Features and layer-wise weighted class-activation maps tailored explicitly for foot segmentation in regional analgesia monitoring. Namely, our contribution is twofold: i) a novel Random Fourier Features layer is introduced to deal with image data to enhance three well-known architectures (FCN, UNet, and ResUNet); ii) three novel quantitive measures are presented to evaluate the interpretability of a given deep learning model devoted to segmentation tasks. Our approach is rigorously evaluated on a demanding dataset of foot thermal images from pregnant women who received epidural anesthesia. Its small size and considerable variability characterize the dataset. Our validation results demonstrate that the proposed methodology not only achieves competitive foot segmentation performance but also significantly enhances the explainability of the process, rendering it well-suited for applications such as epidural insertion during labor.

Interest in drone solutions in forestry applications is growing. Using drones, datasets can be captured flexibly and at high spatial and temporal resolutions when needed. In forestry applications, fundamental tasks include the detection of individual trees, tree species classification, bio-mass estimation, etc. Deep Neural Networks (DNN) have shown superior results when comparing with conventional machine learning methods such as Multi-Layer Perceptron (MLP) in cases of huge input data. The objective of this research was to investigate 3D convolutional neural networks (3D-CNN) to classify three major tree species in a boreal forest: pine, spruce, and birch. The proposed 3D-CNN models were employed to classify tree species in a test site in Finland. The classifiers were trained with a dataset of 3039 manually labelled trees. Then the accuracies were assessed by employing independent datasets of 803 records. To find the most efficient set of feature combination, we compare the performances of 3D-CNN models trained with hyperspectral (HS) channels, RGB channels, and canopy height model (CHM), separately and combined. It is demonstrated that the proposed 3D-CNN model with RGB and HS layers produces the highest classification accuracy. The producer accuracy of the best 3D-CNN classifier on the test dataset were 99.6%, 94.8%, and 97.4% for pines, spruces, and birches, respectively. The best 3D-CNN classifier produced ~5% better classification accuracy than the MLP with all layers. Our results suggest that the proposed method provides excellent classification results with acceptable performance metrics for HS datasets. Our results show that pine class was detectable in most layers. Spruce was most detectable in RGB data, while birch was most detectable in the HS layers. Furthermore, the RGB datasets provide acceptable results for many low-accuracy applications.

Soybean maturity is a trait of critical importance for the development of new soybean cultivars, nevertheless, its characterization based on visual ratings has many challenges. Unmanned aerial vehicles (UAVs) imagery-based high-throughput phenotyping methodologies have been proposed as an alternative to the traditional visual ratings of pod senescence. However, the lack of scalable and accurate methods to extract the desired information from the images remains a significant bottleneck in breeding programs. The objective of this study was to develop an image-based high-throughput phenotyping system for evaluating soybean maturity in breeding programs. Images were acquired twice a week, starting when the earlier lines began maturation until the latest ones were mature. Two complementary convolutional neural networks (CNN) were developed to predict the maturity date. The first using a single date and the second using the five best image dates identified by the first model. The proposed CNN architecture was validated using more than 15,000 ground truth observations from five trials, including data from three growing seasons and two countries. The trained model showed good generalization capability with a root mean squared error lower than two days in four out of five trials. Four methods of estimating prediction uncertainty showed potential at identifying different sources of errors in the maturity date predictions. The architecture used solves limitations of previous research and can be used at scale in commercial breeding programs.

Antimony (Sb) has gained significance as a critical raw material (CRM) within the European Union (EU) due to its strategic importance in various industrial sectors, particularly in the textile industry for flame retardants and as a component of Sb-based semiconductor materials. Moreover, Sb is emerging as a potential alternative for anodes used in lithium-ion batteries, a key element in the Energy transition. This study focused on exploring the feasibility of identifying and quantifying Sb mineralizations through the spectral signature of soils using reflectance spectroscopy, a non-invasive remote sensing technique, and by employing deep learning algorithms such as Convolutional Neural Networks (CNNs). Common signal preprocessing techniques were applied to the spectral data, and the soils were analyzed by Inductively Coupled Plasma Mass Spectrometry (ICP-MS). Despite achieving high R-squared values, the study faces a significant challenge of generalization of the model to new data. Despite the limitations, this study provides valuable insights into potential strategies for future research in this field.

The root system plays an irreplaceable role in plant growth. Its improvement can increase crop productivity. However, such system is still mysterious for us. The underlying mechanism has not been fully uncovered. The investigation on proteins related to the root system is an important means to complete this task. In the previous time, lack of root-related proteins makes it impossible to adopt machine learning methods for designing efficient models for the discovery of novel root-related proteins. Recently, a public database on root-related proteins was set up and machine learning methods can be applied in this field. In this study, we proposed a machine learning based model, named Graph-Root, for identification of root-related proteins. The features derived from protein sequences and one network were extracted, where the former features were processed by graph convolutional neural network and multi-head attention, and the later features abstracted the linkage between proteins. These features were fed into the fully connected layer to make prediction. The 5-fold cross-validation and independent tests suggested its good performance. It also outperformed the only one previous model, SVM-Root. Furthermore, the importance of each feature type and component in the proposed model was investigated.

Skin cancer is one of the widespread diseases that typically develop on the skin due to continuous exposure to sunlight. Although cancer can appear on any part of the human body, skin cancer reports account for over half of all cancer occurrences worldwide. There are substantial obstacles to the precise diagnosis and classification of skin lesions because of the morphological variety and indistinguishable characteristics across skin malignancies. Recently, Deep Learning models have been used in the field of image-based lesion diagnosis, and it has demonstrated diagnostic efficiency on par with that of dermatologists. To increase classification efficiency and accuracy for skin lesions, a cutting-edge multi-layer deep Convolutional Neural Network (CNN) termed SkinLesNet has been built in this study. The ResNetV50 and VGG16 models have been carefully compared to review the performance of the proposed model. The dataset used in this study, PAD-UFES-20, contains 1314 samples in total and includes three common forms of skin lesions. The proposed approach, SkinLesNet, significantly outperforms the well-known compared models in the given conditions

Barrett's esopaghagus (BE) is a known precursor of esophageal adenocarcinoma (EAC). Patients with BE undergo regular surveillance to early detect stages of EAC. Volumetric laser endomicroscopy (VLE) is a novel technology capable of imaging the inner tissue layers of the esophagus over a 6-cm length scan. However, interpretation of full VLE scans is still a challenge for human observers. In this work, we train an ensemble of deep convolutional neural networks to detect neoplasia in BE patients, using a dataset of images acquired with VLE in a multicenter study. We achieve values of AUC=$0.96$ on the unseen test dataset and we compare our results with previous work done with VLE analysis. Our method for detecting neoplasia in BE patients facilitates future advances on patient treatment and provides clinicians with new assisting solutions to process and better understand VLE data.

Parkinson’s Disease (PD) is a neurodegenerative and progressive disease that impacts the nerve cells in the brain and varies from person to person. The exact cause of PD is still unknown, and the diagnosis of PD does not include a specific objective test with certainty. Although deep learning has made great progress in medical neuroimaging analysis, these methods are very susceptible to biases present in neuroimaging datasets. An innovative decorrelated deep learning technique is introduced to mitigate class bias and scanner bias while simultaneously focusing on finding distinguishing characteristics in resting-state functional MRI (rs-fMRI) data, which assist in recognizing the PD with good accuracy. The decorrelation function reduces the non-linear correlation between features and bias in order to learn bias-invariant features. The Parkinson’s Progression Markers Initiative (PPMI) dataset, referred to as a single scanner imbalanced dataset in this study used to validate our method. The imbalanced dataset problem affects the performance of the deep learning framework by overfitting to the majority class. To resolve this problem, we propose a new Decorrelated Convolutional Neural Networks (DcCNN) framework by applying decorrelation-based optimization to Convolutional Neural Networks(CNN). An analysis of evaluation metrics comparisons shows that integrating the decorrelation function boosts the performance of PD recognition by removing class bias. Specifically, our DcCNN model performs significantly better than existing traditional approaches to tackle the imbalance problem. Finally, the same framework can be extended to create scanner invariant features without significantly impacting the performance of a model. The obtained dataset is a multi-scanner dataset which leads to scanner bias due to the differences in acquisition protocols and scanners. The multi-scanner dataset is a combination of two datasets, namely PPMI and FTLDNI - frontotemporal lobar degeneration neuroimaging initiative (NIFD) dataset. The results of t-distributed stochastic neighbor embedding (t-SNE) and scanner classification accuracy of our proposed Feature Extraction-DcCNN (FE-DcCNN) model validated the effective removal of scanner bias. Our method achieves an average accuracy of 77.80% on a multi-scanner dataset for differentiating PD from healthy control, which is superior to the DcCNN model trained on a single scanner imbalanced dataset.

Object segmentation on remotely-sensed images: aerial (or very high resolution, VHS) images and satellite (or high resolution, HR) images, has been applied to many application domains, especially road extraction in which the segmented objects are served as a mandatory layer in geospatial databases. Several attempts in applying deep convolutional neural network (DCNN) to extract roads from remote sensing images have been made; however, the accuracy is still limited. In this paper, we present an enhanced DCNN framework specifically tailored for road extraction on remote sensing images by applying landscape metrics (LMs) and conditional random fields (CRFs). To improve DCNN, a modern activation function, called exponential linear unit (ELU), is employed in our network resulting in a higher number of and yet more accurate extracted roads. To further reduce falsely classified road objects, a solution based on an adoption of LMs is proposed. Finally, to sharpen the extracted roads, a CRF method is added to our framework. The experiments were conducted on Massachusetts road aerial imagery as well as THEOS satellite imagery data sets. The results showed that our proposed framework outperformed Segnet, the state-of-the-art object segmentation technique on any kinds of remote sensing imagery, in most of the cases in terms of precision, recall, and F1.

In recent times, using image-based deep learning techniques has been very useful for classifying fruits. However, mango is a particular challenge because its shelf life is directly related to the degree of maturity at the time of harvest, which is determined mainly through the visual assessment of the operators when sorting them, doing it incorrectly, coupled with the blows and compressions exposed during storage and transport, trigger the deterioration in their quality, hence the importance of this operation. In this sense, the purpose of this study focuses on designing a computational model based on Deep Learning that uses a convolutional neural network (CNN) parameterized to select mangoes according to their color tone based on their maturity stage. A dataset was established from 201 labeled images, then a CNN architecture was built, obtaining a model on which 3 criteria were evaluated: the number of convolution and grouping layers, the resolution of the input images and the number of training epochs required. As a result, an input image resolution of 32 x 32 pixels was achieved, the process included a convolution layer followed by pooling that fed another layer that processed the results and ended in an output layer with 2 neurons under the One Hot Encoding representation. The model required 20 epochs to perfect its training, achieving 96.04 % accuracy in mango image selection. This study supports the potential of CNNs in quality analysis through mango color selection, even in uncertainty scenarios.

In the current industrial landscape, increasingly pervaded by technological innovations, the adoption of optimized strategies for asset management is becoming a critical key success factor. Among the various strategies available, the “Prognostics and Health Management” strategy is able to support maintenance management decisions more accurately, through continuous monitoring of equipment health and “Remaining Useful Life” forecasting. In the present study, Convolutional Neural Network-based Deep Neural Network techniques are investigated for the Remaining Useful Life prediction of a punch tool, whose degradation is caused by working surface deformations during the machining process. Surface deformation is determined using a 3D scanning sensor capable of returning point clouds with micrometric accuracy during the operation of the punching machine, avoiding both downtime and human intervention. The 3D point clouds thus obtained are transformed into bidimensional image-type maps, i.e., maps of depths and normal vectors, to fully exploit the potential of convolutional neural networks for extracting features. Such maps are then processed by comparing 15 genetically optimized architectures with the transfer learning of 19 pre-trained models, using a classic machine learning approach, i.e., Support Vector Regression, as a benchmark. The achieved results clearly show that, in this specific case, optimized architectures provide performance far superior (MAPE=0.058) to that of transfer learning which, instead, remains at a lower or slightly higher level (MAPE=0.416) than Support Vector Regression (MAPE=0.857).

In this paper, we address the problem of car detection from aerial images using Convolutional Neural Networks (CNN). This problem presents additional challenges as compared to car (or any object) detection from ground images because features of vehicles from aerial images are more difficult to discern. To investigate this issue, we assess the performance of two state-of-the-art CNN algorithms, namely Faster R-CNN, which is the most popular region-based algorithm, and YOLOv3, which is known to be the fastest detection algorithm. We analyze two datasets with different characteristics to check the impact of various factors, such as UAV's altitude, camera resolution, and object size. The objective of this work is to conduct a robust comparison between these two cutting-edge algorithms. By using a variety of metrics, we show that YOLOv3 yields better performance in most configurations, except that it exhibits a lower recall and less confident detections when object sizes and scales in the testing dataset differ largely from those in the training dataset.

In this paper, we address the problem of car detection from aerial images using Convolutional Neural Networks (CNN). This problem presents additional challenges as compared to car (or any object) detection from ground images because features of vehicles from aerial images are more difficult to discern. To investigate this issue, we assess the performance of two state-of-the-art CNN algorithms, namely Faster R-CNN, which is the most popular region-based algorithm, and YOLOv3, which is known to be the fastest detection algorithm. We analyze two datasets with different characteristics to check the impact of various factors, such as UAV’s altitude, camera resolution, and object size. The objective of this work is to conduct a robust comparison between these two cutting-edge algorithms. By using a variety of metrics, we show that none of the two algorithms outperforms the other in all cases.

Popular geo-computer vision works make use of aerial imager with sizes ranging from 64 × 64 to 1024 × 1024 pixels without any overlap, although the learning process of deep learning models can be affected by the reduced semantic context or the lack of information near image boundaries. In this work, the impact of three image sizes (256 × 256, 512 × 512, and 1024 × 1024 pixels) and two image overlap levels (no overlap and 12.5% overlap) on the performance of road classification models was statistically evaluated. For this, two convolutional neural networks used in various tasks of geospatial object extraction were trained (using the same hyperparameters) on a large dataset (containing aerial image data covering 8650 km² of the Spanish territory that was labelled with binary road information) under twelve different scenarios, each scenario featuring a different combination of tile size and overlap. To assess their generalisation capacity, the performance of all resulting models was evaluated on a data from novel areas covering approximately 825 km². The performance metrics obtained were analysed using appropriate descriptive and inferential statistical techniques to evaluate the impact of the performance at distinct levels of the fixed factors (tile size, tile overlap, neural network architecture). Statistical tests were applied to study the main and interaction effects of the fixed factors on the performance. A significance level of 0.05 was applied to all the null hypothesis tests. The results were highly significant for the main effects (p-values lower than 0.001), while the two-way and three-way interaction effects among them had different levels of significance. The results indicate that the training of road classification models on images with a higher tile size (more semantic context) and tile overlap (additional border context and continuity) significantly impacts their performance. The best model was trained on a dataset featuring tiles with a size of 1024 × 1024 pixels and a 12.5% overlap and achieved a loss value of 0.0984, an F1 score of 0.8728, and an ROC-AUC score of 0.9766, together with an error rate of 3.5% on the test set.

One of the critical tools for early detection and subsequent evaluation of the incidence of lung diseases is chest radiography. This study presents a real-world implementation of a convolutional neural network (CNN) based Carebot Covid app to detect COVID-19 from chest X-ray (CXR) images. Our proposed model takes the form of a simple and intuitive application. Used CNN can be deployed as a STOW-RS prediction endpoint for direct implementation into DICOM viewers. The results of this study show that the deep learning model based on DenseNet and ResNet architecture can detect SARS-CoV-2 from CXR images with precision of 0.981, recall of 0.962 and AP of 0.993.

In the remote sensing domain, it is crucial to complete semantic segmentation on the raster images, e.g., river, building, forest, etc, on raster images. A deep convolutional encoder--decoder (DCED) network is the state-of-the-art semantic segmentation method for remotely sensed images. However, the accuracy is still limited, since the network is not designed for remotely sensed images and the training data in this domain is deficient. In this paper, we aim to propose a novel CNN for semantic segmentation particularly for remote sensing corpora with three main contributions. First, we propose applying a recent CNN called a global convolutional network (GCN), since it can capture different resolutions by extracting multi-scale features from different stages of the network. Additionally, we further enhance the network by improving its backbone using larger numbers of layers, which is suitable for medium resolution remotely sensed images. Second, "channel attention'' is presented in our network in order to select the most discriminative filters (features). Third, "domain-specific transfer learning'' is introduced to alleviate the scarcity issue by utilizing other remotely sensed corpora with different resolutions as pre-trained data. The experiment was then conducted on two given datasets: (i) medium resolution data collected from Landsat-8 satellite and (ii) very high resolution data called the ISPRS Vaihingen Challenge Dataset. The results show that our networks outperformed DCED in terms of $F1$ for 17.48% and 2.49% on medium and very high resolution corpora, respectively.

Deep Learning (DL) has made significant advances in computer vision with the advent of Vision Transformers (ViT). Unlike Convolutional Neural Networks (CNNs), ViTs use self-attention to extract both local and global features from image data, and then use residual connections to feed these features directly into a fully networked multilayer perceptron head. In hospitals, hematologists prepare peripheral blood smears (PBSs) and read them under a medical microscope to detect abnormalities in blood counts such as leukemia. However, this task is time-consuming and prone to human error. This study investigates the transfer learning process of Google ViT and ImageNet CNNs to automate the reading of PBSs. The study used two online PBS datasets, PBC and BCCD, and transferred them into balanced datasets to investigate the influence of data amount and noise immunity on both neural networks. The PBC results show that Google ViT is an excellent DL neural solution for data scarcity. The BCCD results show that Google ViT is superior to ImageNet CNNs in dealing with unclean, noisy image data because it is able to extract both global and local features and use residual connections, despite the additional time and computational overhead.

In recent times, with the increase of Artificial Neural Network (ANN), deep learning has brought a dramatic twist in the field of machine learning by making it more Artificial Intelligence (AI). Deep learning is used remarkably used in vast ranges of fields because of its diverse range of applications such as surveillance, health, medicine, sports, robotics, drones etc. In deep learning, Convolutional Neural Network (CNN) is at the center of spectacular advances that mixes Artificial Neural Network (ANN) and up to date deep learning strategies. It has been used broadly in pattern recognition, sentence classification, speech recognition, face recognition, text categorization, document analysis, scene, and handwritten digit recognition. The goal of this paper is to observe the variation of accuracies of CNN to classify handwritten digits using various numbers of hidden layer and epochs and to make the comparison between the accuracies. For this performance evaluation of CNN, we performed our experiment using Modified National Institute of Standards and Technology (MNIST) dataset. Further, the network is trained using stochastic gradient descent and the backpropagation algorithm.

Reliable prediction of infectious disease is an essential need for public health organizations to minimize or prevent the disease's spread. With the increase of data growth in the healthcare sector, accurate analysis of such data helps for early disease detection and better patient care. With the availability of huge data, it is a great deal for predicting and handling an epidemic occurrence. To control the epidemic disease, a new Artificial Intelligence called Manhattan Hessian Linear Feature Embedding Based Tucker Congruence Deep Convolutional Network (MHLFE-TCDCN) is introduced with higher accuracy and minimum time consumption. The MIoT device is used to collect the data from the patient body. The proposed MHLFE-TCDCN includes two major processes namely feature selection and classification with multiple layers such as input, two hidden layers, and one output layer. First, the feature selection process is performed using the Manhattan Hessian Locally Linear Embedding technique in the first hidden layer to minimize the dimensionality and time complexity. After selecting the feature, the classification is performed using the Tucker coefficient of congruence regression. The Tucker coefficient of congruence measures the similarity between the training and testing disease data. A positive similarity level is used to accurately identify the disease at the output layer based on the similarity measure. Experimental assessment is carried out with various parameters such as accuracy, precision, recall, F-measure, and computational time concerning several numbers of patient data. The quantitative results and discussion that the proposed MHLFE-TCDCN technique achieves higher accuracy with minimal computation time as compared to the conventional methods.

: Submerged aquatic vegetation (SAV) are highly efficient at carbon sequestration and, despite their relatively small distribution globally, are recognized as a potentially valuable component of climate change mitigation. However, SAV mapping in tidal marshes presents a challenge due to optically complex constituents in the water. The emergence and advancement of deep learn-ing-based techniques in the field of habitat mapping with remote sensing imagery provides an opportunity to address this challenge. In this study, an analytical framework was developed to quantify the carbon sequestration of SAV habitats in the Atchafalaya River Delta Estuary from field and remote sensing observations using deep convolutional neural network (DCNN) tech-niques. A U-Net based model, Wetland-SAV Network, was trained to identify SAV percent cover (high, medium, and low) as well as other estuarine habitat types from Landsat 8/9-OLI data. The areal extent of SAV was up to 8% of the total area (47,000 ha) with a significant loss of SAV habitats observed post-Hurricane Barry (~2,300 ha) in 2019. The habitat areas and habitat-specific carbon fluxes were then used to quantify net greenhouse gas (GHG) flux of the study area for with/without SAV scenarios in a Carbon Balance Model. The total net GHG flux was in the range of -0.13 ± 0.06 to -0.86 ± 0.37 ×105 tonne CO2e yr-1 and increased up to 40% (-0.23 ± 0.10 to -0.90 ± 0.39 ×105 tonne CO2e yr-1 ) when SAV was accounted for within the calculation. At the hectare scale, inclusion of SAV resulted in an increase of ~60% for net GHG sink in shallow areas adjacent to emergent marsh where SAV was abundant. This is the first attempt at remotely mapping SAV in coastal Louisiana as well as a first quantification of net GHG flux at the scale of hectares to thousands of hectares, accounting for SAV within these sub-tropical coastal delta marshes. Remote sensing and deep learning models have high potential for mapping and monitoring of SAV in turbid sub-tropical coastal deltas as a component of increasing accuracy of net GHG flux estimates at small (hectare) and large (coastal basin) scales.

Online and offline shopping trip have different impacts on various aspects of urban life, such as e-commerce, transportation systems, and sustainability. Therefore, it is important to evaluate the factors that influence their choices. We use a hybrid machine learning model that combines a gray wolf optimization algorithm and a deep convolutional neural network to estimate shopping trip based on a survey of 1,000 active e-commerce users who made successful orders in both online and offline services in the last 20 days of 2021 in areas 2 and 5 of Tehran. The gray wolf optimization algorithm performs feature selection and hyperparameter tuning for the deep convolutional neural network, which is a powerful deep learning model for image recognition and classification. The results show that our model achieves an accuracy of 97.81% with an MSE of 0.325 by selecting seven out of ten features. The most important features are delivery cost, delivery time, product price, car ownership. In addition, comparing the performance of the proposed method with other methods showed that the proposed algorithm with an accuracy of 97.81%, the accuracies of the single deep learning model, MLP neural network, decision tree, and KNN models were 95.63%, 90.0%, 86.49%, and 80.16%, respectively.

Deep learning is an emerging area of machine learning (ML) research. It comprises multiple hidden layers of artificial neural networks. The deep learn- ing methodology applies nonlinear transformations and model abstractions of high level in large databases. The recent advancements in deep learning architec- tures within numerous fields have already provided significant contributions in artificial intelligence. This article presents a state of the art survey on the contri- butions and the novel applications of deep learning. The following review chron- ologically presents how and in what major applications deep learning algorithms have been utilized. Furthermore, the superior and beneficial of the deep learning methodology and its hierarchy in layers and nonlinear operations are presented and compared with the more conventional algorithms in the common applica- tions. The state of the art survey further provides a general overview on the novel concept and the ever-increasing advantages and popularity of deep learning.

"Deep fake" technologies are often associated with face-swapping algorithms based on images or videos. Although much research is focused on developing these techniques, there is still room for improvement, especially in human and visual assessment. In this study, we present an innovative approach to the problem of face replacement based on one model while preserving background features. Our model uses the Adaptive Attentional Denormalization method from FaceShifter while integrating identity features from the ArcFace and BiSeNet models to support the attribute extraction process. We used Fast GAN as a generative model to optimize and accelerate the training process on smaller data sets. Our research proves the effectiveness of the proposed approach by combining innovative methods and architectures in one coherent solution.

The popularity of deep reinforcement learning (DRL) methods in economics have been exponentially increased. DRL through a wide range of capabilities from reinforcement learning (RL) and deep learning (DL) for handling sophisticated dynamic business environments offers vast opportunities. DRL is characterized by scalability with the potential to be applied to high-dimensional problems in conjunction with noisy and nonlinear patterns of economic data. In this work, we first consider a brief review of DL, RL, and deep RL methods in diverse applications in economics providing an in-depth insight into the state of the art. Furthermore, the architecture of DRL applied to economic applications is investigated in order to highlight the complexity, robustness, accuracy, performance, computational tasks, risk constraints, and profitability. The survey results indicate that DRL can provide better performance and higher accuracy as compared to the traditional algorithms while facing real economic problems at the presence of risk parameters and the ever-increasing uncertainties.

Diabetes mellitus is a popular life-threatening disease and patients may gradually have started suffering from other diabetes-causing diseases such as heart attacks, stroke, hypertension, blurry vision, blindness, foot ulcer, amputation, kidney damage and other organ failures before diagnosis. Early detection can help reduce the fatality of this disease. Deep learning models have proven very useful in disease detection and computer-aided diagnosis. In this work, we proposed a deep unsupervised machine learning model for early detection of diabetes using voting ensemble feature selection and deep belief neural networks (DBN). Dataset was obtained from an online repository containing responses of prediagnosed patients to direct questionnaires administered in Sylhet Diabetes Hospital in Sylhet, Bangladesh. The dataset was preprocessed and preprocessed. Features were reduced using the ensemble feature selector. The DBN model was pretrained and tuned to obtain optimal performance. The model was also compared with other models with no multiple hidden layers. The DBN performed at its relative best with F1-measure, precision and recall of 1.00, 0.92 and 1.00 respectively. We conclude that DBN is a useful tool for an unsupervised early prediction of Type II diabetes mellitus.

We propose a deep-learning approach for the joint MIMO detection and channel decoding problem. Conventional MIMO receivers adopt a model-based approach for MIMO detection and channel decoding in linear or iterative manners. However, due to the complex MIMO signal model, the optimal solution to the joint MIMO detection and channel decoding problem (i.e., the maximum likelihood decoding of the transmitted codewords from the received MIMO signals) is computationally infeasible. As a practical measure, the current model-based MIMO receivers all use suboptimal MIMO decoding methods with affordable computational complexities. This work applies the latest advances in deep learning for the design of MIMO receivers. In particular, we leverage deep neural networks (DNN) with supervised training to solve the joint MIMO detection and channel decoding problem. We show that DNN can be trained to give much better decoding performance than conventional MIMO receivers do. Our simulations show that a DNN implementation consisting of seven hidden layers can outperform conventional model-based linear or iterative receivers. This performance improvement points to a new direction for future MIMO receiver design.

Deep networks-based hashing has gained significant popularity in recent years, particularly in the field of image retrieval. However, most existing methods only focus on extracting semantic information from the final layer, disregarding valuable structural information that contains important semantic details crucial for effective hash learning. To address this limitation and improve image retrieval accuracy, we propose a novel deep hashing method called Deep Supervised Hashing by Fusing Multiscale Deep Features (DSHFMDF). Our approach involves extracting multiscale features from multiple convolutional layers and fusing them to generate more robust representations for efficient image retrieval. Experimental results on CIFAR10 and NUS-WIDE datasets demonstrate that our method surpasses the performance of state-of-the-art hashing techniques.

Groundwater is the most extracted raw material in the world with global annual withdrawal rates of 800–1500 km3/year. In East and Southern Africa, 70 % of the population are reliant on shallow groundwater as their primary drinking water source. With increased population growth, intensification of agriculture and industrialization, conflicting demands on groundwater present a challenge to achieve the Sustainable Development Goals (6,3,11,12,15). Between 2015 and 2018, the Horn of Africa was affected by a series of climatic induced events, namely El Nino, La Nina and the Indian Ocean Diopole. These events modified the variability of rainfall patterns and resulted in long periods of low rainfall, low recharge and high evapotranspiration. As a result, shallow aquifers in alluvial deposits of Somali region have low yields and produce brackish and saline water. That situation prompted humanitarian water professionals to finance the transportation of water from selected locations with high groundwater potential through water trucks to areas facing groundwater depletion and drought. To address this challenge, UNICEF explored alternative, sustainable deeper groundwater sources that could be extracted using solar water pumping technology for multi water use. This paper describes a three-phase methodology of deep groundwater development of wells in the Ogaden Jesoma sandstone aquifers of the Somali region of the Horn of Africa to a depth of 600 meters below ground level. The results concluded that the deep sandstone aquifer of Jesoma can provide fresh water with yields of 15 l/s to the local population of Somali region. to the study provided insights into deep groundwater identification and development as well as adaptive deep boreholes drilling as a source for climate resilient water supplies.

Artificial intelligence (AI) is a current trend in computer science, which extends itself its amazing capacities to other technologies such as mechatronics and robotics. Going beyond technological applications, the philosophy behind AI is that there is a vague and potential convergence of artificial manufacture and natural world although the limiting approach may be still very far away, but why? The implicit problem is that Darwin theory of evolution focuses on natural world where breeding conservation is the cornerstone of the existence of creature world but there is no similar concept of breeding conservation in artificial world whose things are created by human. However, after developing for a long time until now, AI issues an interesting concept of generation in which artifacts created by computer science can derive their new generations inheriting their aspects / characteristics. Such generated artifacts make us look back on offsprings by the process of breeding conservation in natural world. Therefore, it is possible to think that AI generation, which is a recent subject of AI, is a significant development in computer science as well as high-tech domain. AI generation does not help us to reach near biological evolution even in the case that AI can combine with biological technology but, AI generation can help us to extend our viewpoint about Darwin theory of evolution as well as there may exist some uncertain relationship between man-made world and natural world. Anyhow AI generation is a current important subject in AI and there are two main generative models in computer science: 1) generative model that applies large language model into generating natural language texts understandable by human and 2) generative model that applies deep neural network into generating digital content such as sound, image, and video. This technical report focuses on deep generative model (DGM) for digital content generation, which is a short summary of approaches to implement DGMs. Researchers can read this work as an introduction to DGM with easily understandable explanations.

The rise of the Internet has enabled people to express their opinions on various subjects through social media, blogs, and website comments. As a result, there has been a significant increase in research on sentiment analysis. However, most of the research efforts have focused on analyzing sentiment in English-language data, neglecting the wealth of information available in other languages. In this paper, we provide a comprehensive review of the current state-of-the-art in multilingual sentiment analysis. The survey investigates techniques for data preprocessing, representation learning, and feature extraction in multilingual sentiment analysis. It explores cross-lingual transfer learning, domain adaptation, and data augmentation methods that enhance the performance of sentiment analysis models, particularly in low-resource languages and domains. It provides insights into the state-of-the-art approaches, challenges, and opportunities in this evolving field and encourages further advancements in multilingual sentiment analysis research.

When training deep neural networks tasks, the most popular choices are cross-entropy loss. On the other hand, in general speaking, a decent loss function can take on shapes that are considerably more flexible and ought to be adapted for different activities and datasets. In most of the classification tasks, generally if the true class is not correctly recognized by the network (top1), that class is placed among the five classes with the highest probability (top5). This shows that the network does not necessarily recognize the correct class with a low probability, but a class similar to it (such as 3 vs. 8 in mnist) assigns a higher probability and this causes a mistake in that task. Accordingly, we proposed a loss function deals with the error of class that the neural network incorrectly recognized as correct, in addition to the correct class error. We call our proposed loss as False Positive Loss (FPL), with the intention of viewing and designing loss functions not only through the utilization of true class but also through the utilization of the value of false positive classes. One of the core properties of our proposed loss is full adaptability, which makes False Positive Loss be fully capable of getting reformulated by using other widely used loss functions formulas based on the task or the need of the users. Extensive experimental results demonstrate that our suggested loss function outperforms other well-known losses on a variety of tasks and datasets. As can be observed, the performance of our False Positive Loss is superior to that of the cross-entropy loss when it comes to tasks involving 2D picture classification. We have compared our loss with cross entropy as the most common classification loss function on some models (such as ResNet-18, ResNet-50 and Efficientnet-V2) through classification known as a basic computer vision task. with both random or pre-trained initial weights. As a result, in some cases the models with our loss outperform the same tasks with cross entropy from the viewpoint of some metric (i.e. accuracy and FP). For example, the resnet-50 on cifar-10 dataset with random initialization indicated a top1 accuracy of 94.93 with cross entropy and 95.25 with our loss, while for top5 accuracy the results are 99.86 and 99.87, respectively.

We study the brain segmentation by dividing the brain into multiple tissues. Given possible brain segmentation by deep, machine learning can be efficiently exploited to expedite the segmentation process in the clinical practice. To accomplish segmentation process, a MRI and tissues transfer using generative adversarial networks is proposed. Given the better result, we propose the transfer model using GAN. For the case of the brain tissues, white matter (WM), gray matter (GM) and cerebrospinal fluid (CSF) are segmented. Empirical results show that this proposed model significantly improved segmentation results compared to the stat-of-the-art results. Furthermore, a dice coefficient (DC) metric is used to evaluate the model performance.

This paper compares a series of traditional and deep learning methodologies for the segmentation of textures. Six well-known texture composites first published by Randen and Hus{\o}y were used to compare traditional segmentation techniques (co-occurrence, filtering, local binary patterns, watershed, multiresolution sub-band filtering) against a deep-learning approach based on the U-Net architecture. For the latter, the effects of depth of the network, number of epochs and different optimisation algorithms were investigated. Overall, the best results were provided by the deep-learning approach. However, the best results were distributed within the parameters, and many configurations provided results well below the traditional techniques.

In this paper we propose a computational framework that merges mechanistic modeling with deep neural networks obeying the Systems Biology Markup Language (SBML) standard. Over the last 20 years, the systems biology community has developed a large number of mechanistic models in SBML that are currently stored in public databases. With the proposed framework, existing SBML mechanistic models may be upgraded to hybrid systems through the incorporation of deep neural networks into the model core, using a freely available python tool. The so-formed hybrid mechanistic/neural network models are trained with a deep learning algorithm based on the adaptive moment estimation method (ADAM), stochastic regularization and semidirect sensitivity equations. The trained hybrid models are encoded in SBML and stored back in model databases, where they can be further analyzed as regular SBML models. The application of this approach is illustrated with three well-known case studies: the threonine synthesis model in Escherichia coli, the P58IPK signal transduction model, and the Yeast glycolytic oscillations model. The proposed framework is expected to greatly facilitate the widespread use of hybrid modeling techniques for systems biology applications.

Electricity load forecasting is a crucial undertaking within all the deregulated markets globally. In contemporary times, the transition from conventional electricity grids to Smart Grids constitutes an area where extensive research is conducted on a global scale. Among the research challenges, the investigation of Deep Transfer Learning (DTL) in the field of electricity load forecasting represents a fundamental effort that imparts generality to Artificial Intelligence applications, due to new capabilities, such as knowledge transfer and computational power reduction. In this paper a comprehensive study is conducted for day-ahead electricity load forecasting. For this purpose, three Sequence-to-Sequence (Seq2seq) Deep Learning (DL) models are used, namely the Multilayer Perceptron (MLP), the Convolutional Neural Network (CNN) and the Ensemble Learning Model (ELM), which is consisted of the weighted combination of the outputs of MLP and CNN models. Also, the study focuses on the development of different forecasting strategies based on DTL and emphasizing the way the datasets are trained and fine-tuning for higher forecasting accuracy. In order to implement the forecasting strategies using Deep Learning models, load datasets from three Greek islands, Rhodes, Lesvos, and Chios, are used. The main purpose is to apply DTL for day-ahead predictions (1-24 hours) for each month of the year for Chios dataset after training and fine-tuning the models using the datasets of the three islands in various combinations. After several trials, four DTL strategies are illustrated. In the first strategy (DTL Case 1), each of the three DL models is trained using only Lesvos dataset, while fine-tuning is performed on the dataset of Chios island, in order to create day-ahead predictions for Chios load. In the second strategy (DTL Case 2), data from both Lesvos and Rhodes concurrently are used for the DL model training period, and fine-tuning is performed on the data from Chios. The third DTL strategy (DTL Case 3) contains the training of the DL models using Lesvos dataset, and the testing period performed directly on the Chios dataset without fine-tuning. The fourth strategy is a Multi-task Deep Learning (MTDL) approach, which has been extensively studied in recent years. In MTDL, the three DL models are trained simultaneously on all three datasets and the final predictions are made on the unknown part of the dataset of Chios. In this paper, we explore the performance of DTL and compare the results with those produced with MTDL. The results demonstrated that DTL can be applied with high efficiency for day-ahead load forecasting. Specifically, the two cases with fine-tuning (DTL Case 1 and 2) outperformed MTDL in terms of load prediction accuracy. Regarding the DL models, all three exhibit very high prediction accuracy, especially in the two cases with fine-tuning. The ELM excels compared to the single models. More specifically, for conducting day-ahead predictions, it has been concluded that the MLP model presents best monthly forecasts with a MAPE of 6.24% and 6.01% for the first two cases, the CNN model presents best monthly forecasts with a MAPE of 5.57% and 5.60% respectively and the ELM model achieves best monthly forecasts with a MAPE of 5.29% and 5.31%, respectively, indicating the very high accuracy it can achieve.

The extensive utilization of drones has led to numerous scenarios that encompass both advantageous and perilous outcomes. By using deep learning techniques, it is aimed to reduce the dangerous effects of drone use through early detection of drones.The state of the art of the study is evaluation of deep learning approaches such as pre trained YOLOv8 drone detection for security issues. In this study, dataset collected by Mehdi Özel for a UAV competition are sourced from GitHub. These images are labeled using Roboflow, and the model is trained on Google Colab. To enhance model performance, dataset augmentation techniques including rotation and blurring are implemented. Following these steps, a significant precision value of 0.946, a notable recall value of 0.9605 and a considerable precision-recall curve value of 0.978 are achieved, surpassing many popular models such as Mask CNN, CNN, and YOLOv5.

The study aimed to investigate the impact of incorporating deep learning strategies in mathematics instruction on students' problem-solving abilities, critical thinking skills, academic achievement, and mathematical skills development. The study population consisted of 280 students, with a sample size of 162 participants. Employing a purely quantitative approach, data collection utilized questionnaires and achievement tests as instruments. Data analysis involved simple linear regression, descriptive analysis, and Pearson correlation. The results of the study revealed that there was no significant enhancement observed in students' problem-solving abilities and critical thinking skills when deep learning strategies were incorporated into mathematics instruction. Additionally, students' attitudes towards deep learning activities in mathematics showed no effect on their academic achievement. Furthermore, students' engagement levels in deep learning activities did not impact their development of mathematical skills. Lastly, the frequency of deep learning activities in the mathematics classroom did not show a significant relationship with students' academic performance. These findings suggest that the implementation of deep learning strategies in mathematics instruction did not yield notable improvements in the specified areas based on the study outcomes.

To identify objects in images, a complex set of skills is needed that includes understanding the context and being able to determine the borders of objects. In computer vision, this task is known as semantic segmentation and it involves categorizing each pixel in an image. It is crucial in many real-world situations: for autonomous vehicles, it enables the identification of objects in the surrounding area; in medical diagnosis, it enhances the ability to detect dangerous pathologies early, thereby reducing the risk of serious consequences. In this study, we compare the performance of various ensembles of convolutional and transformer neural networks. Ensembles can be created, e.g, by varying the loss function, data augmentation method or the learning rate strategy. Our proposed ensemble, which is based on the simple average rule, demonstrates exceptional performance on several datasets. All the resources used in this study are available online at the following GitHub repository: https://github.com/LorisNanni.

Polar Mesospheric Summer Echoes (PMSEs) are radar echoes that are observed in the mesosphere during the Arctic summer months in the polar regions. By studying PMSE, researchers can gain insights into physical and chemical processes that occur in the upper atmosphere specifically in the 80 to 90 km altitude range. In this paper, we employ fully convolutional networks such as UNET and UNET++ for the purposes of segmenting PMSE from the EISCAT VHF dataset. First, experiments are performed to find suitable weights and hyperparameters for UNET and UNET++. Second, different loss functions are tested to find one suitable for our task. Third, as the number of PMSE samples used is relatively small that can lead to poor generalization. To address this, image-level and object-level augmentation methods are employed. Four, we briefly explain our findings by employing layerwise relevance propagation.

In the field of low power wireless networks, one of the techniques on which many researchers are putting their efforts is related to positioning methodologies such as fingerprinting in dense urban areas. This paper presents an experimental study aimed at quantifying the mean location estimation error in densely urbanized areas.Using a dataset made available by the University of Antwerp, a neural network was implemented with the aim of providing the position of the end-devices. In this way it was possible to measure the mean location estimation error in an area with high urban density. The results obtained show an accuracy in the localization of the end-device of less than 150 meters.This result would make it possible to use the fingerprint instead of alternative, energy consuming, methodologies such as GPS in IoT (Internet of Things) applications where battery life is the primary requirement to be met.

Deep learning, specifically the supervised approach, has proved to be a breakthrough in depth prediction. However, the generalization ability of deep networks is still limited, and they cannot maintain a satisfactory performance on some inputs. Addressing a similar problem in the segmentation field, a scheme (f-BRS) has been proposed to refine predictions in the inference time. f-BRS adapts an intermediate activation function to each input by using user clicks as sparse labels. Given the similarity between user clicks and sparse depth maps, this paper aims at extending the application of f-BRS to depth prediction. Our experiments show that f-BRS, fused with a depth estimation baseline, is trapped in local optima, and fails to improve the network predictions. To resolve that, we propose a double-stage adaptive refinement scheme (DARS). In the first stage, a Delaunay-based correction module significantly improves the depth generated by a baseline network. In the second stage, a particle swarm optimizer (PSO) delineates the estimation through fine-tuning f-BRS parameters—that is, scales and biases. DARS is evaluated on an outdoor benchmark, KITTI, and an indoor benchmark, NYUv2 while for both the network is pre-trained on KITTI. The proposed scheme outperformed rival methods on both datasets.

Numerous approaches address over-fitting in neural networks: by imposing a penalty on the parameters of the network (L1, L2, etc); by changing the network stochastically (drop-out, Gaussian noise, etc.); or by transforming the input data (batch normalization, etc.). In contrast, we aim to ensure that a minimum amount of supporting evidence is present when fitting the model parameters to the training data. This, at the single neuron level, is equivalent to ensuring that both sides of the separating hyperplane (for a standard artificial neuron) have a minimum number of data points — noting that these points need not belong to the same class for the inner layers. We firstly benchmark the results of this approach on the standard Fashion-MINST dataset, comparing it to various regularization techniques. Interestingly, we note that by nudging each neuron to divide, at least in part, its input data, the resulting networks make use of each neuron, avoiding a hyperplane completely on one side of its input data (which is equivalent to a constant into the next layers). To illustrate this point, we study the prevalence of saturated nodes throughout training, showing that neurons are activated more frequently and earlier in training when using this regularization approach. A direct consequence of the improved neuron activation is that deep networks are now easier to train. This is crucially important when the network topology is not known a priori and fitting often remains stuck in a suboptimal local minima. We demonstrate this property by training a network of increasing depth (and constant width): most regularization approaches will result in increasingly frequent training failures (over different random seeds) whilst the proposed evidence-based regularization significantly outperforms in its ability to train deep networks.

Fracture analysis of reinforced concrete deep beam strengthened with carbon fiber-reinforced polymer (CFRP) plates was carried out. The present research aims to find out whether the crack propagation in a strengthened deep beam follows linear elastic fracture mechanics (LEFM) theory or nonlinear fracture mechanics theory. To do so, a new energy release rate based on nonlinear fracture mechanics theory was formulated on the finite element method and the discrete cohesive zone model (DCZM) was developed in deep beams. To validate and compare with numerical models, three deep beams with rectangular cross-sections were tested. The code results based on nonlinear fracture mechanics models were compared with experimental results and ABAQUS results carried out based on LEFM. The predicted values of initial stiffness, yielding point and failure load, energy absorption, and compressive strain in the concrete obtained by the proposed model were very close to the experimental results. However, the ABAQUS software results have greater differences with the experimental results. For example, the predicted failure load for the shear-strengthened deep beam using the proposed model has only 6.3% differences compared to the experimental result. However, the predicted failure load using ABAQUS software based on LEFM has greater differences (25.1%) compared to the experimental result.