Performance Of Convolutional Neural Networks Research Articles

A Comparative Analysis of Convolutional Neural Networks for American Sign Language Recognition

Abstract. Sign language recognition is an important technology that makes it possible for ordinary people to be able to communicate with deaf people, fostering inclusivity and accessibility. The came out of deep learning technology has completely changed the field by enabling the automatic extraction and learning of hierarchical features from raw data, leading to significant improvements in recognition accuracy. This paper presents a comprehensive comparative analysis of different Convolutional Neural Network (CNN) architectures for recognizing American Sign Language (ASL) signs. Utilizing the sign language dataset, which contains 24 classes of ASL letters represented by 28x28 grayscale images, the author evaluated the performance of a Basic CNN, a Modified Residual Network (ResNet)-50, and a LeNet-5 model. This study emphasizes the impact of architectural choices on recognition accuracy and computational efficiency. Results indicate that while ResNet-50 demonstrates superior accuracy, fluctuating significantly during initial training, the Basic CNN and LeNet-5 models offer greater stability with slightly lower accuracy. This work concludes that despite the initial challenges, deep learning models, particularly ResNet-50, show promise for ASL recognition, highlighting the need for diverse and enriched datasets to improve model reliability in real-world scenarios.

Analysis of athletes’ technical action based on deep learning

In order to raise athletes’ technical proficiency and overall performance, this paper uses deep learning technology to precisely assess table tennis technical activities. The paper builds a hybrid neural network model for technical action analysis of table tennis players, based on Convolutional Neural Network (CNN) and Long Short-Term Memory (LSTM) in deep learning. First, CNN extracts the spatial characteristics from the video frames. After that, an LSTM processes these data in a time series to accurately recognize and classify the movements of athletes. Lastly, through trials, the model is trained and evaluated using the Table Tennis Tracking Network (TTNET) dataset. The experimental findings demonstrate that: (1) In table tennis technical action analysis, the suggested deep learning model performs better than other comparative models. The efficacy of combining convolution with sequence learning is fully demonstrated by the CNN-LSTM hybrid model, which performs best in all indicators when compared to the CNN, LSTM, Multi-Objective Function (MOF), and Convolutional Neural Network—Long Short-Term Memory (CNN-LSTM) combined models. Its accuracy rate, precision rate, recall rate, and F1 score are 0.923, 0.918, 0.925, and 0.921, respectively. In contrast, the performances of LSTM and CNN are also excellent, but the performance of MOF model is relatively low; (2) In the classification of technical actions, the model has the highest classification accuracy of service, reaching 0.930, and the classification accuracy of other technical actions such as forehand stroke, backhand stroke and spike is also above 0.9, and the time sequence consistency index also maintains a high level, indicating that the model can effectively identify and analyze table tennis technical actions. In addition, the performance evaluation of real-time feedback shows that the model can achieve low processing time and feedback delay in video data processing with different lengths, which ensures the real-time and reliability in practical applications. These results show that the proposed model can not only provide accurate technical action recognition, but also provide timely and effective feedback in practical application, which has high practical value. The results of this paper prove the potential of deep learning technology in the analysis of athletes’ technical actions, and provide scientific basis and effective tools for the technical training and optimization of table tennis players.

Exploring spatial reasoning performances of CNN on linear layout dataset

Abstract Spatial reasoning, a fundamental aspect of human intelligence, is essential for machine learning models to understand and interpret object relationships. It is crucial for numerous real-world applications, ranging from autonomous navigation to urban planning. The lack of comprehensive datasets limits the development and evaluation of models that can effectively handle spatial reasoning tasks. Existing datasets often contain complex spatial reasoning problems with overlapping spatial relationships, making it challenging to diagnose specific aspects that a model struggles with. We address this gap by introducing a new dataset of linear layouts. This dataset is systematically designed to exhibit a range of spatial relations and complexity levels. Analyzing spatial reasoning through linear layout generation offers a more structured and manageable approach to understanding how models learn and interpret spatial relationships. Linear layout generation has broad applicability and is of fundamental importance in design and optimization. To benchmark dataset, we develop LinLayCNN, a generic data-driven method that applies shallow, one-dimensional convolutional neural network (CNN), to generate linear layouts in an iterative process. Experimental results reveal that LinLayCNN can effectively solve fundamental spatial challenges even with the relatively small size of the training set. It is capable of precise object placement, making it a robust tool for linear layout generation. Current layout generation methods focus on domain-specific solutions, often fail to maintain the precision needed for technical domains, such as accurate sizing, and object counting. They also require a substantial amount of data to function effectively. LinLayCNN overcame these issues. This study also enhances our understanding of CNNs' capabilities in spatial reasoning, highlight their potential to advance the field of layout generation. As a result, our approach establishes a clear benchmark for evaluating spatial reasoning and aids in development of models that can more effectively understand and reason about space.

Deep learning approaches for bias correction in WRF model outputs for enhanced solar and wind energy estimation: A case study in East and West Malaysia

Accurate estimation of wind and solar energy potentials is crucial for successfully integrating renewable energy into power grids. Traditional numerical weather prediction models, such as the Weather Research and Forecasting (WRF) model, often suffer from biases that lead to inaccurate energy forecasts. This study employs advanced deep learning (DL) techniques to correct these biases in WRF model outputs, specifically to enhance wind and solar energy estimations in East and West Malaysia. Unlike previous studies, this research integrates a diverse array of DL models: Recurrent Neural Networks (RNN), Long Short-Term Memory (LSTM), Convolutional Neural Networks (CNN), and Feedforward Neural Networks (FNN), to address both temporal and spatial prediction challenges. The models were trained and tested using historical weather data and ground-based measurements to improve the accuracy of wind speed and solar radiation predictions. Evaluation metrics, root mean square error (RMSE), mean bias error (MBE), and mean absolute error (MAE), demonstrate the better performance of CNN and FNN models over the sole WRF approach. The findings reveal that CNN achieved the lowest RMSE in wind speed estimation (0.91 in CEMACS and 0.97 in Kuching compared to WRF RMSEs of 1.92 and 1.39). At the same time, FNN significantly improved solar radiation prediction (RMSE of 86.86 in Kuching and 99.23 in CEMACS compared to WRF RMSEs of 154.44 and 370.66). Given the low wind speeds, the corrected data from CNN was used to estimate wind energy at 536 kWh at Kuching and 0 kWh at CEMACS. FNN-corrected data was also used to estimate solar energy at 19 kWh and 18 kWh at Kuching and CEMACS, respectively. This research not only shows the effectiveness of DL in mitigating biases in numerical weather prediction models but also contributes a novel methodology for reliable renewable energy assessments.

Data augmentation via warping transforms for modeling natural variability in the corneal endothelium enhances semi-supervised segmentation.

Image segmentation of the corneal endothelium with deep convolutional neural networks (CNN) is challenging due to the scarcity of expert-annotated data. This work proposes a data augmentation technique via warping to enhance the performance of semi-supervised training of CNNs for accurate segmentation. We use a unique augmentation process for images and masks involving keypoint extraction, Delaunay triangulation, local affine transformations, and mask refinement. This approach accurately captures the natural variability of the corneal endothelium, enriching the dataset with realistic and diverse images. The proposed method achieved an increase in the mean intersection over union (mIoU) and Dice coefficient (DC) metrics of 17.2% and 4.8% respectively, for the segmentation task in corneal endothelial images on multiple CNN architectures. Our data augmentation strategy successfully models the natural variability in corneal endothelial images, thereby enhancing the performance and generalization capabilities of semi-supervised CNNs in medical image cell segmentation tasks.

Convolutional Neural Networks: A Comprehensive Evaluation and Benchmarking of Pooling Layer Variants

Convolutional Neural Networks (CNNs) are a class of deep neural networks that have proven highly effective in areas such as image and video recognition. CNNs typically include several types of layers, such as convolutional layers, activation layers, pooling layers, and fully connected layers, all of which contribute to the network’s ability to recognize patterns and features. The pooling layer, which often follows the convolutional layer, is crucial for reducing computational complexity by performing down-sampling while maintaining essential features. This layer’s role in balancing the symmetry of information across the network is vital for optimal performance. However, the choice of pooling method is often based on intuition, which can lead to less accurate or efficient results. This research compares various standard pooling methods (MAX and AVERAGE pooling) on standard datasets (MNIST, CIFAR-10, and CIFAR-100) to determine the most effective approach in preserving detail, performance, and overall computational efficiency while maintaining the symmetry necessary for robust CNN performance.

Combining Biology-based and MRI Data-driven Modeling to Predict Response to Neoadjuvant Chemotherapy in Patients with Triple-Negative Breast Cancer.

"Just Accepted" papers have undergone full peer review and have been accepted for publication in Radiology: Artificial Intelligence. This article will undergo copyediting, layout, and proof review before it is published in its final version. Please note that during production of the final copyedited article, errors may be discovered which could affect the content. Purpose To combine deep learning and biology-based modeling to predict the response of locally advanced, triple negative breast cancer before initiating neoadjuvant chemotherapy (NAC). Materials and Methods In this retrospective study, a biology-based mathematical model of tumor response to NAC was constructed and calibrated on a patient-specific basis using imaging data from patients enrolled in the MD Anderson ARTEMIS trial (ClinicalTrials.gov, NCT02276443) between April 2018 and May 2021. To relate the calibrated parameters in the biology-based model and pretreatment MRI data, a convolutional neural network (CNN) was employed. The CNN predictions of the calibrated model parameters were used to estimate tumor response at the end of NAC. CNN performance in the estimations of total tumor volume (TTV), total tumor cellularity (TTC), and tumor status was evaluated. Model-predicted TTC and TTV measurements were compared with MRI-based measurements using the concordance correlation coefficient (CCC), and area under the receiver operating characteristic curve (for predicting pathologic complete response at the end of NAC). Results The study included 118 female patients (median age, 51 [range, 29-78] years). For comparison of CNN predicted to measured change in TTC and TTV over the course of NAC, the CCCs were 0.95 (95% CI: 0.90-0.98) and 0.94 (95% CI: 0.87-0.97), respectively. CNN-predicted TTC and TTV had an AUC of 0.72 (95% CI: 0.34-0.94) and 0.72 (95% CI: 0.40-0.95) for predicting tumor status at the time of surgery, respectively. Conclusion Deep learning integrated with a biology-based mathematical model showed good performance in predicting the spatial and temporal evolution of a patient's tumor during NAC using only pre-NAC MRI data. ©RSNA, 2024.

Intelligent Fault Diagnosis Method Based on Neural Network Compression for Rolling Bearings

Rolling bearings are often exposed to high speeds and pressures, leading to the symmetry in their rotating structure being disrupted, which can lead to serious failures. Intelligent rolling bearing fault diagnosis is a critical part of ensuring operation of machinery, and it has been facilitated by the growing popularity of convolutional neural networks (CNNs). The outstanding performance of fault diagnosis CNNs results from complex and redundant network structures and parameters, resulting in huge storage and computational requirements, which makes it challenging to implement these models in resource-limited industrial devices. This study aims to address this problem by proposing a comprehensive compression method for CNNs that is applied to intelligent fault diagnosis. It involves several different compression methods, including tensor train decomposition, parameter quantization, and knowledge distillation for deep network compression. This results in a significant decrease in redundancy and speeding up the training of CNN models. Firstly, tensor train decomposition is applied to reduce redundant connections in both convolutional and fully connected layers. The next step is to perform parameter quantization to minimize the bits needed for parameter representation and storage. Finally, knowledge distillation is used to restore accuracy to the compressed model. The effectiveness of the proposed approach is confirmed by an experiment and ablation study with different models on several datasets. The results show that it can significantly reduce redundant information and floating-point operations with little degradation in accuracy. Notably, on the CWRU dataset, with about 60% parameter reduction, there is no degradation in our model’s accuracy. The proposed approach is a new attempt at the intelligent fault diagnosis of rolling bearings in industrial equipment.

AI-powered IoT and UAV systems for real-time detection and prevention of illegal logging

Illegal logging poses a significant threat to environmental conservation, contributing to deforestation, ecosystem disruption, and economic losses. Traditional detection methods, such as ground patrols and satellite imagery, are often ineffective, expensive, and time-consuming. This paper introduces a novel Unmanned Aerial Vehicles (UAV)-based IoT system for chainsaw noise detection and prevention. The system uses low-cost IoT nodes with microphones and microcontrollers to detect chainsaw sounds via a Convolutional Neural Network (CNN) model. These nodes reduce power consumption, allowing the battery to last at least 8 years. Detected chainsaw sounds are transmitted to a LoRaWAN gateway and forwarded to a cloud server. The cloud server processes the data and dispatches a drone equipped with a camera to the site. The drone streams live video to the cloud, where AI algorithms analyze the footage for human presence. If detected, the drone activates sound and light alarms to deter illegal logging. Experimental results show that IoT nodes can detect chainsaw sounds up to 100 m, with the CNN model achieving an accuracy of 99.37 % and a loss of 0.0185. Comparisons with Logistic Regression, Decision Tree, Random Forest, and SVM demonstrated the superior performance of CNN. This UAV-IoT system offers a fast, reliable, scalable, and real-time solution for detecting and preventing illegal logging. Leveraging advanced AI and IoT technologies, this system enhances environmental monitoring and conservation efforts. It represents a significant step forward in the fight against illegal logging, providing a practical, effective, and sustainable solution.

Automated Recognition of Submerged Body-like Objects in Sonar Images Using Convolutional Neural Networks

The Police Robot for Inspection and Mapping of Underwater Evidence (PRIME) is an uncrewed surface vehicle (USV) currently being developed for underwater search and recovery teams to assist in crime scene investigation. The USV maps underwater scenes using sidescan sonar (SSS). Test exercises use a clothed mannequin lying on the seafloor as a target object to evaluate system performance. A robust, automated method for detecting human body-shaped objects is required to maximise operational functionality. The use of a convolutional neural network (CNN) for automatic target recognition (ATR) is proposed. SSS image data acquired from four different locations during previous missions were used to build a dataset consisting of two classes, i.e., a binary classification problem. The target object class consisted of 166 196 × 196 pixel image snippets of the underwater mannequin, whereas the non-target class consisted of 13,054 examples. Due to the large class imbalance in the dataset, CNN models were trained with six different imbalance ratios. Two different pre-trained models (ResNet-50 and Xception) were compared, and trained via transfer learning. This paper presents results from the CNNs and details the training methods used. Larger datasets are shown to improve CNN performance despite class imbalance, achieving average F1 scores of 97% in image classification. Average F1 scores for target vs background classification with unseen data are only 47% but the end result is enhanced by combining multiple weak classification results in an ensemble average. The combined output, represented as a georeferenced heatmap, accurately indicates the target object location with a high detection confidence and one false positive of low confidence. The CNN approach shows improved object detection performance when compared to the currently used ATR method.

Transfer learning classification of suspicious lesions on breast ultrasound: is there room to avoid biopsies of benign lesions?

BackgroundBreast cancer (BC) is the most common malignancy in women and the second cause of cancer death. In recent years, there has been a strong development in artificial intelligence (AI) applications in medical imaging for several tasks. Our aim was to evaluate the potential of transfer learning with convolutional neural networks (CNNs) in discriminating suspicious breast lesions on ultrasound images.MethodsTransfer learning performances of five different CNNs (Inception V3, Xception, Densenet121, VGG 16, and ResNet50) were evaluated on a public and on an institutional dataset (526 and 392 images, respectively), customizing the top layers for the specific task. Institutional images were contoured by an expert radiologist and processed to feed the CNNs for training and testing. Postimaging biopsies were used as a reference standard for classification. The area under the receiver operating curve (AUROC) was used to assess diagnostic performance.ResultsNetworks performed very well on the public dataset (AUROC 0.938–0.996). The direct generalization to the institutional dataset resulted in lower performances (max AUROC 0.676); however, when tested on BI-RADS 3 and BI-RADS 5 only, results were improved (max AUROC 0.792). Good results were achieved on the institutional dataset (AUROC 0.759–0.818) and, when selecting a threshold of 2% for classification, a sensitivity of 0.983 was obtained for three of five CNNs, with the potential to spare biopsy in 15.3%–18.6% of patients.ConclusionIn conclusion, transfer learning with CNNs may achieve high sensitivity and might be used as a support tool in managing suspicious breast lesions on ultrasound images.Relevance statementTransfer learning is a powerful technique to exploit the performances of well-trained CNNs for image classification. In a clinical scenario, it might be useful for the management of suspicious breast lesions on breast ultrasound, potentially sparing biopsy in a non-negligible number of patients.Key PointsProperly trained CNNs with transfer learning are highly effective in differentiating benign and malignant lesions on breast ultrasound.Setting clinical thresholds increased sensitivity.CNNs might be useful as support tools in managing suspicious lesions on breast ultrasound.Graphical

Artificial intelligence derived electrocardiograms: a comparative study of network architectures predicting sex and left ventricular dysfunction

Abstract Background Within cardiovascular medicine, AI-driven electrocardiogram (ECG) analysis has emerged as a powerful tool, capable of predicting clinically significant abnormalities. The increasing growth of wearable technologies for ECG monitoring underscores the importance of advancing AI methodologies. While Convolutional Neural Networks (CNNs) have traditionally been the go-to choice for ECG analysis, the rise of Vision Transformers (ViTs), a novel computer vision paradigm, raises the question of their potential superiority. Purpose To compare the performance of CNNs and ViTs in ECG analysis, and to assess the efficiency of lean models using truncated ECG recordings. Methods We trained a model using 12-lead ECGs as input for the binary classification of two outputs: Sex (Male or Female) and Left ventricular dysfunction (LVD, defined as below 35%), determined by echocardiography within 2 weeks of the ECG. Models were developed using full standard ECGs, single lead and truncated recordings (1,2 and 4 10 seconds). We compared CNNs (PyTorch) and ViTs (HuggingFace) performance. We analyzed the explainability of our findings, representing the AI model's focus of interest on ECGs from normal cases versus those with reduced ejection fraction, thereby validating its diagnostic discernment (Fig 1) Results We identified 150,691 and 29,422 patients with valid ECG tests for the Sex and LVD prediction models, respectively. For the Sex outcome, the AUROCs were 0.911 (95% CI: 0.908-0.914) and 0.898 (95% CI: 0.894-0.901) for the CNN-based model and the ViT-based models, respectively, with the difference statistically significant (P-Value < 0.001). For the LVD outcome, the AUROCs were comparable at 0.878 (95% CI: 0.864-0.892) and 0.866 (95% CI: 0.853-0.879) for the CNN-based model and the ViT-based models, respectively, (P-Value = 0.056). Furthermore, we found that 98.8% for the maximal predictive power of the models was achieved within a 2 second time frame of a 12 lead ECG. Similarly, a single lead based model achieved a relatively high AUC for the prediction of LVD. Conclusions AI-ECG models developed with ViT architecture did not surpass CNN-based models in sex classification and LVD identification. Our study highlights the significance of lean models allowing for rapid prediction within 2-second tracings or using a single lead, demonstrates the potential for streamlined AI-ECG applications in clinical practice.

ENHANCING CNN PERFORMANCE WITH RIGHT-NEIGHBOR DEVIATION (RND) POOLING: A FOCUS ON MULTI-CLASS BRAIN TUMOR CLASSIFICATION IN MRI IMAGES

In recent years, the field of image processing and computer vision has been deeply transformed by the advent of Convolutional Neural Network (CNN). These deep learning models, inspired by biological processes, have demonstrated remarkable success in tasks such as image classification, object detection, and semantic segmentation. A critical component of this success is feature extraction, where the pooling layer within CNN serves as a nonlinear down-sampling technique. This layer reduces the spatial size of the convolved feature map through operations like average or maximum pooling, significantly impacting the model’s performance in complex tasks. This research introduces a novel pooling algorithm, Right-Neighbor Deviation (RND), designed to enhance the performance of CNN. The RND pooling method aims to capture detailed spatial features and hierarchical representations more effectively, thereby improving the accuracy and resilience of deep learning models. A dataset of multi-class brain tumor classification, incorporating MRI images from the BR35H dataset, is utilized to explore the potential benefits of RND pooling compared to traditional methods. The effectiveness of the proposed method is evaluated using performance metrics including accuracy, precision, recall, and F1 score, achieving generalized values of 98.86%, 98.84%, 98.80%, and 98.80%, respectively. The results highlight the potential of RND pooling to significantly advance the state of the art in brain tumor classification using CNNs, potentially leading to improved diagnosis and treatment outcomes.

Automatic Compressive Sensing of Shack–Hartmann Sensors Based on the Vision Transformer

Shack–Hartmann wavefront sensors (SHWFSs) are crucial for detecting distortions in adaptive optics systems, but the accuracy of wavefront reconstruction is often hampered by low guide star brightness or strong atmospheric turbulence. This study introduces a new method of using the Vision Transformer model to process image information from SHWFSs. Compared with previous traditional methods, this model can assign a weight value to each subaperture by considering the position and image information of each subaperture of this sensor, and it can process to obtain wavefront reconstruction results. Comparative evaluations using simulated SHWFS light intensity images and corresponding deformable mirror command vectors demonstrate the robustness and accuracy of the Vision Transformer under various guide star magnitudes and atmospheric conditions, compared to convolutional neural networks (CNNs), represented in this study by Residual Neural Network (ResNet), which are widely used by other scholars. Notably, normalization preprocessing significantly improves the CNN performance (improving Strehl ratio by up to 0.2 under low turbulence) while having a varied impact on the Vision Transformer, improving its performance under a low turbulence intensity and high brightness (Strehl ratio up to 0.8) but deteriorating under a high turbulence intensity and low brightness (Strehl ratio reduced to about 0.05). Overall, the Vision Transformer consistently outperforms CNN models across all tested conditions, enhancing the Strehl ratio by an average of 0.2 more than CNNs.

Tumor detection in breast cancer pathology patches using a Multi-scale Multi-head Self-attention Ensemble Network on Whole Slide Images

Breast cancer (BC) is the most common type of cancer among women globally and is one of the leading causes of cancer-related deaths among women. In the diagnosis of BC, histopathological assessment is the gold standard, where automated tumor detection technologies play a pivotal role. Utilizing Convolutional Neural Networks (CNNs) for automated analysis of image patches from Whole Slide Images (WSIs) enhances detection accuracy and alleviates the workload of pathologists. However, CNNs often face limitations in handling pathological patches due to a lack of sufficient contextual information and limited feature generation capabilities. To address this, we propose a novel Multi-scale Multi-head Self-attention Ensemble Network (MMSEN), which integrates a multi-scale feature generation module, a convolutional self-attention module, and an adaptive feature integration with an output module, effectively optimizing the performance of classical CNNs. The design of MMSEN optimizes the capture of key information and the comprehensive integration of features in WSIs pathological patches, significantly enhancing the precision of tumor detection. Validation results from a five-fold cross-validation experiment on the PatchCamelyon (PCam) dataset demonstrate that MMSEN achieves a ROC-AUC of 99.01% ± 0.02%, an F1-score of 98.00% ± 0.08%, a Balanced Accuracy (B-Acc) of 98.00% ± 0.08%, and a Matthews Correlation Coefficient (MCC) of 96.00% ± 0.16% (p<0.05). These results demonstrate the effectiveness and potential of MMSEN in detecting tumors from pathological patches in WSIs for BC.

Improving acoustic species identification using data augmentation within a deep learning framework

Convolutional neural networks (CNNs) are effective tools for acoustic classification tasks such as species identification. Large datasets of labelled recordings are required to develop CNN classifiers which can be difficult to obtain, particularly if species are rare or vocalise infrequently. Additionally, data often requires manual labelling which can be time consuming requiring expert analysis. Artificially generating data using augmentation can address these challenges, however the impact of data augmentation on CNN performance is poorly understood and often omitted in bioacoustic studies. Here, we empirically test the impact of CNN architecture and 20 data augmentation methods on classifier performance. We use acoustic identification of 18 small mammal species as a case study of a species group that can be effectively surveyed by acoustic monitoring, but recordings for training data are scarce and difficult to collect. Networks that achieved the highest accuracy across all sample sizes was a 10-layer CNN (96.43 %) and a pre-trained ResNet50 model (96.37 %). Overall, all augmentation effects improved ResNet50 model performance and 17 effects improved Conv10 performance, increasing relative change in accuracy (RCA) by 0.021–0.641. Three augmentation effects negatively impacted Conv10 RCA by −0.042 to −0.182. We also show that adding augmented data when the number of original samples is low has the greatest positive impact on accuracy and this effect was larger with ResNet50 models. Our work demonstrates that using data augmentation where few original samples are available can considerably improve model performance and highlights the potential of augmentation in developing acoustic classifiers for species where data are limited or difficult to obtain.

Lamb mode identification based on lightweight CNN

ABSTRACT In this study, a lightweight convolutional neural network (CNN) is employed to identify Lamb modes. The proposed approach consists of five convolutional and pooling layers, then a fully-connected layer and a sigmoid layer. In which, the first convolutional layer is a wide-scale kernel. Lamb wave responses based on froward modelling are obtained for different plate materials (aluminium, steel and titanium), different excitation frequencies (250 kHz, 500 kHz), and different excitation cycles (4-cycle, 5-cycle). 16800 Lamb wave samples labelled by ‘A0 mode’ and ‘S0 mode’ are beforehand and hosted in a database, then trained via the lightweight CNN. In validation process, the lightweight CNN reaches 100% accuracy. The performance of light-weight CNN is also compared with some popular networks. Now, the well-trained network can be used to identify Lamb mode. Some responses are stimulated by ABAQUS under different excitation signal, different propagating distance, different plate material, and the predicted results via the lightweight CNN are all right. In addition, the extensibility of the network is validated by identifying new-converted Lamb mode correctly.

VMCNN: Integrating Vedic Mathematics for Enhanced Convolutional Neural Network Performance in Image Classification Tasks

In the rapidly advancing field of deep learning, Convolutional Neural Networks (CNNs) have emerged as a cornerstone for various image classification tasks. However, the existing CNN architectures and optimization methods often grapple with trade-offs between computational efficiency, accuracy, and generalizability. This work addresses these limitations by integrating principles of Vedic Mathematics, an ancient Indian mathematical system known for its efficiency and simplicity, into CNN optimization. Traditional optimization techniques often fall short in balancing the computational load with the accuracy of the model, particularly in diverse and complex datasets. To overcome these challenges, we propose a novel model that incorporates specific Vedic Mathematical Sutras, namely Urdhva-Tiryakbhyam, Anurupyena, Nikhilam Navatashcaramam Dashatah, Shunyam Saamyasamuccaye, Paravartya Yojayet, Sankalana-vyavakalanabhyam, Ekadhikina Purvena, and Gunitasamuchyah, for optimizing internal operations of CNNs. These Sutras were meticulously selected for their potential to simplify computational processes, enhance parallel processing capabilities, and optimize training algorithms. The application of these sutras has led to a remarkable improvement in the performance of CNNs across various datasets, including ImageNet, CIFAR, ChestXRay8, and Architectural Heritage Datasets. The optimized CNN models demonstrated a significant enhancement in classification accuracy (9.5% increase), precision (8.3% increase), recall (8.5% increase), area under the curve (AUC) (4.9% increase), MAE (5.5% improvement), and reduction in delay (6.5% decrease) compared to existing methods. The integration of Vedic Mathematics into CNN optimization not only paves the way for more efficient and accurate image classification models but also opens new avenues for interdisciplinary research, blending ancient mathematical wisdom with contemporary artificial intelligence techniques. This advancement has profound implications for various applications, including medical imaging, autonomous vehicles, and heritage conservation, thereby contributing significantly to the field of AI and computational efficiency.

Bimodal biometric recognition system using Convolutional Neural Networks and fusion of deep audiovisual feature vectors

In recent years, interest has grown in the use biometric systems for identity authentication tasks in digital services, forensic and security applications. A unimodal system (employing a single biometric trait) with high performance is still vulnerable to falsification attacks such as spoofing. For this reason, research on multimodal biometrics (employing various biometric traits) has increased to reinforce security, increase recognition performance, and make false identity authentication more difficult. In this paper, we propose a bimodal system that combines speech and face modalities by concatenating their feature vectors, these vectors are extracted from two convolutional neural networks (CNN) and used for identity verification. The performance of unimodal CNNs was evaluated individually and compared to the bimodal system of concatenated vectors. A data augmentation scheme is used for both modalities to evaluate different operation conditions. Results were measured in terms of Equal Error Rate (EER).

Network Robustness Prediction: Influence of Training Data Distributions.

Network robustness refers to the ability of a network to continue its functioning against malicious attacks, which is critical for various natural and industrial networks. Network robustness can be quantitatively measured by a sequence of values that record the remaining functionality after a sequential node- or edge-removal attacks. Robustness evaluations are traditionally determined by attack simulations, which are computationally very time-consuming and sometimes practically infeasible. The convolutional neural network (CNN)-based prediction provides a cost-efficient approach to fast evaluating the network robustness. In this article, the prediction performances of the learning feature representation-based CNN (LFR-CNN) and PATCHY-SAN methods are compared through extensively empirical experiments. Specifically, three distributions of network size in the training data are investigated, including the uniform, Gaussian, and extra distributions. The relationship between the CNN input size and the dimension of the evaluated network is studied. Extensive experimental results reveal that compared to the training data of uniform distribution, the Gaussian and extra distributions can significantly improve both the prediction performance and the generalizability, for both LFR-CNN and PATCHY-SAN, and for various functionality robustness. The extension ability of LFR-CNN is significantly better than PATCHY-SAN, verified by extensive comparisons on predicting the robustness of unseen networks. In general, LFR-CNN outperforms PATCHY-SAN, and thus LFR-CNN is recommended over PATCHY-SAN. However, since both LFR-CNN and PATCHY-SAN have advantages for different scenarios, the optimal settings of the input size of CNN are recommended under different configurations.