Global Average Pooling Layer Research Articles

Generative Adversarial Network Based Music Generation

Music has been an integral part of human civilization personally and culturally. Historically, music has been generated using various instruments, or natural sounds like water drops, or unconventional musical instruments like metal or glass-wares. At present, technologies like Musical Instrument Digital Interface (MIDI) are used to generate music electronically. This research investigates the use of Generative Adversarial Networks (GANs) for beginner-friendly music production. This model uses Long Short-Term Memory (LSTM) generator and Patch GAN as discriminator for the GAN architecture. The generator consists of input layer, embedding layer, LSTM layer and generates output with a SoftMax function. Similarly, the discriminator consists of a convolution layer, the output of which is averaged by the global average pooling layer, and output is generated by the sigmoid function. The model is trained on Maestro MIDI dataset. We make the process understandable by delving into the implementation specifics and outlining the fundamental concepts of music. Our effective model highlights the potential of GANs in music composition by producing cohesive music. After training for 50 epochs, the model exhibited a remarkable precision of 91.82 percent. This project uses the combination of Artificial Intelligence (AI) with music theory to provide intriguing new opportunities in the field of music. The model can be beneficial for different industries like gaming, music, entertainment, education, etc.

Class activation map-based slicing-concatenation and contrastive learning: A novel strategy for record-level atrial fibrillation detection

BackgroundDeep learning-based models for atrial fibrillation (AF) detection require extensive training data, which often necessitates labor-intensive professional annotation. While data augmentation techniques have been employed to mitigate the scarcity of annotated electrocardiogram (ECG) data, specific augmentation methods tailored for recording-level ECG annotations are lacking. This gap hampers the development of robust deep learning models for AF detection. MethodsWe propose a novel strategy, a combination of Class Activation Map-based Slicing-Concatenation (CAM-SC) data augmentation and contrastive learning, to address the current challenges. Initially, a baseline model incorporating a global average pooling layer is trained for classification and to generate class activation maps (CAMs), which highlight indicative ECG segments. After that, in each recording, indicative and non-indicative segments are sliced. These segments are subsequently concatenated randomly based on starting and ending Q points of QRS complexes, with indicative segments preserved to maintain label correctness. Finally, the augmented dataset undergoes contrastive learning to learn general representations, thereby enhancing AF detection performance. ResultsUsing ResNet-101 as the baseline model, training with the augmented data yielded the highest F1-score of 0.861 on the Computing in Cardiology (CinC) Challenge 2017 dataset, a typical AF dataset with recording-level annotations. The metrics outperform most previous studies. ConclusionsThis study introduces an innovative data augmentation method specifically designed for recording-level ECG annotations, significantly enhancing AF detection using deep learning models. This approach has substantial implications for future AF detection research.

Feature-Level Fusion Multi-Sensor Aggregation Temporal Network for Smartphone-Based Human Activity Recognition

Smartphone-based Human Activity Recognition (HAR) identifies human movements using inertial signals gathered from multiple smartphone sensors. Generally, these signals are stacked as one (data-level fusion) and fed into deep learning algorithms for feature extractions. This research studies feature-level fusion, individually processing inertial signals from each sensor, and proposes a lightweight deep temporal learning model, Feature-Level Fusion Multi-Sensor Aggregation Temporal Network (FLF-MSATN), that performs feature extraction on inertial signals from each sensor separately. The raw signals, segmented into equally sized time windows, are passed into individual Dilated-Pooled Convolutional Heads (DPC Heads) for temporal feature analysis. Each DPC Head has a spatiotemporal block containing dilated causal convolutions and average pooling, to extract underlying patterns. The DPC Heads’ outputs are concatenated and passed into a Global Average Pooling layer to generate a condensed confidence map before activity classification. FLF-MSATN is assessed using a subject-independent protocol on a publicly available HAR dataset, UCI HAR, and a self-collected HAR dataset, achieving 96.67% and 82.70% accuracies, respectively. A Data-Level Fusion MSATN is built to compare and verify the model performance attained by the proposed FLF-MSATN. The empirical results show that implementing FLF-MSATN enhances the accuracy by ~3.4% for UCI HAR and ~9.68% for self-collected datasets.

Research on Fault Diagnosis of Rolling Bearing Based on Gramian Angular Field and Lightweight Model.

Due to the limitations of deep learning models in processing one-dimensional signal feature extraction, and high model complexity leading to low training accuracy and large consumption of computing resources, this paper innovatively proposes a rolling bearing fault diagnosis method based on Gramian Angular Field (GAF) and enhanced lightweight residual network. Firstly, the one-dimensional signal is transformed into a two-dimensional GAF image, fully preserving the signal's temporal dependency. Secondly, to address the parameter redundancy and high computational complexity of the ResNet-18 model, its residual blocks are improved. The second convolutional layer in the downsampling residual blocks is removed, traditional convolutional layers are replaced with depthwise separable convolutions, and the lightweight Efficient Channel Attention (ECA) module is embedded after each residual block. This further enhances the model's ability to capture key features while maintaining low computational cost, resulting in a lightweight model referred to as E-ResNet13. Finally, the generated GAF feature maps are fed into the E-ResNet13 model for training, and through a global average pooling layer, they are mapped to a fully connected layer for classifying the faults of rolling bearings. Verifying the superiority of the proposed GAF-E-ResNet13 model, experimental results show that the GAF image encoding method achieves higher fault recognition accuracy compared to other encoding methods. Compared with other intelligent diagnosis methods, the E-ResNet13 model demonstrates strong diagnostic performance and generalization capability under both a single condition and complex varying conditions, fully proving the innovation and practicality of this method.

Modified deep inductive transfer learning diagnostic systems for diabetic retinopathy severity levels classification

Diabetic Retinopathy (DR), a retinal illness that degenerates the retina and causes blindness, can be effectively treated with early detection and examination. Although expensive and unpleasant, manual retinography is the gold standard for DR diagnosis. Many Deep Learning (DL)-based algorithms have shown promise as deep learning (DR) diagnostic tools, performing similarly to human picture evaluation. These strategies work best with fine-tuned hyperparameters and huge databases. To develop Deep Inductive Transfer Learning-based Diagnostic Systems (DITL-DS) for Multiclass DR Severity Classification. The benchmark Indian Diabetic Retinopathy Image Dataset (IDRiD) is chosen first. Next, the dataset is pre-processed by proportionally scaling the image to 128*128 and using Convolutional Local Area Histogram Equalization (CLAHE). Third, the uneven class label dataset is balanced for efficiency. Fourth, a global average pooling layer and bottlenecking are added to five DITL models—Inception V3, ResNet34, EfficientNet B0, VGG16, and Xception—to compensate for information loss. Finally, we compare our improved approaches to five basic DITL classifiers. Our modified Xception model separates severity stages best, with an accuracy of 0.9840, precision of 0.9845, recall of 0.9650, AUC of 0.9979, and AUC-ROC of 0.9902. Precision, recall, and F1 scores are 0.97, 0.99, and 0.98 for stage0, 0.99, 1, and 1 for stage1, 0.98, 0.94, and 0.96 for stage2, 0.99, 0.99, and 0.99 for stage3, and 1, 1, and 1 for stage4. This may help clinicians objectively diagnose DR early. To ensure robustness, the model’s generalizability was validated using APTOS (Asia Pacific Tele-Ophthalmology Society).

Corrosion image classification method based on EfficientNetV2

Corrosion is one of the key factors leading to material failure, which can occur in facilities and equipment closely related to people's lives, causing structural damage and thus affecting the safety of people's lives and property. To identify corrosion more effectively across multiple facilities and equipment, this paper utilizes a corrosion binary classification dataset containing various materials to develop a CNN classification model for better detection and distinction of material corrosion, using a methodological paradigm of transfer learning and fine-tuning. The proposed model implementation initially uses data augmentation to enhance the dataset and employs different sizes of EfficientNetV2 for training, evaluated using Confusion Matrix, ROC curve, and the values of Precision, Recall, and F1-score. To further enhance the testing results, this paper focuses on the impact of using the Global Average Pooling layer versus the Global Max Pooling layer, as well as the number of fine-tuning layers. The results show that the Global Average Pooling layer performs better, and EfficientNetV2B0 with a fine-tuning rate of 20 %, and EfficientNetV2S with a fine-tuning rate of 15 %, achieve the highest testing accuracy of 0.9176, an ROC-AUC value of 0.97, and Precision, Recall, and F1-Score values exceeding 0.9. These findings can be served as a reference for other corrosion classification models which uses EfficientNetV2.

Detection of rice pests based on RepMob-Yolo

Abstract Aiming at the disadvantages of high labor intensity, poor real-time performance, and low efficiency of traditional methods for identifying rice pests, a rice pest identification model based on RepMob-Yolo is proposed. MobileNetV3 is used as the backbone network, and the RepMob structure is used to replace the conventional downsampling structure in MobileNetV3, the GELU is used to replace the ReLU and the last four convolutional layers of MobileNetV3, the global average pooling layer and the fully connected layer are pruned to realize the lightweight design of the model. The DANet module is embedded to realize the extraction of effective information by the model, extract the semantic information of higher key features, and improve the recognition accuracy. The FPN structure is combined with the RepMob structure for feature extraction and prediction, which improves the operation speed of the model. The experimental results show that RepMob-Yolo achieves 97.68% accuracy, 96.00% recall, 96.82% prediction rate, and 146.37 f/s frame rate in the identification of public pest datasets, which provides technical support for accurate and real-time identification of rice pests.

Identification and diagnosis of schizophrenia based on multichannel EEG and CNN deep learning model

This paper proposes a high-accuracy EEG-based schizophrenia (SZ) detection approach. Unlike comparable literature studies employing conventional machine learning algorithms, our method autonomously extracts the necessary features for network training from EEG recordings. The proposed model is a ten-layered CNN that contains a max pooling layer, a Global Average Pooling layer, four convolution layers, two dropout layers for overfitting prevention, and two fully connected layers. The efficiency of the suggested method was assessed using the ten-fold-cross validation technique and the EEG records of 14 healthy subjects and 14 SZ patients. The obtained mean accuracy score was 99.18 %. To confirm the high mean accuracy attained, we tested the model on unseen data with a near-perfect accuracy score (almost 100 %). In addition, the results we obtained outperform numerous other comparable works.

A Robust Deep Learning Framework Based on Spectrograms for Heart Sound Classification.

Heart sound analysis plays an important role in early detecting heart disease. However, manual detection requires doctors with extensive clinical experience, which increases uncertainty for the task, especially in medically underdeveloped areas. This paper proposes a robust neural network structure with an improved attention module for automatic classification of heart sound wave. In the preprocessing stage, noise removal with Butterworth bandpass filter is first adopted, and then heart sound recordings are converted into time-frequency spectrum by short-time Fourier transform (STFT). The model is driven by STFT spectrum. It automatically extracts features through four down sample blocks with different filters. Subsequently, an improved attention module based on Squeeze-and-Excitation module and coordinate attention module is developed for feature fusion. Finally, the neural network will give a category for heart sound waves based on the learned features. The global average pooling layer is adopted for reducing the model's weight and avoiding overfitting, while focal loss is further introduced as the loss function to minimize the data imbalance problem. Validation experiments have been conducted on two publicly available datasets, and the results well demonstrate the effectiveness and advantages of our method.

Operational risk assessment of hydropower units based on PSSCA-VMD-CNN-GBiLSTM and multi-feature fusion

Hydroelectric units are an essential part of hydropower plants nowadays, and assessing and predicting the operating conditions of hydroelectric units is of great significance for the reliable operation of hydropower plants. Firstly, an optimization method based on polynomial regression with Pearson and Spearman fusion and Sine Cosine Algorithm (SCA) algorithm (PSSCA) is proposed to optimize the parameters in the variational modal decomposition (VMD), so as to suppress the influence of noise on the vibration signals of hydroelectric units. Then, the sample entropy and root-mean-square fusion of the VMD decomposed modal functions are reconstructed into a hybrid feature vector set so that this vector set can maximally contain the feature information of the vibration signals of the hydroelectric units. Finally, the CNN-GBiLSTM assessment model is constructed by combining the Convolutional Neural Network (CNN) model, the Bi-directional Long Short-Term Memory (BiLSTM) model, and the global average pooling layer, which is used to predict the upper and lower limits of the healthy vibration of the hydropower unit, and to construct the risk indicator of the hydropower unit operation by combining the values of the upper and lower limits. The results of simulation and comparative experiments show that the method proposed in this study can visually represent the three operational risk scenarios of a hydropower unit: low risk, medium risk, and high risk. It can effectively accomplish the operational risk assessment of hydropower units and improve the assessment accuracy and assessment speed, thus proving the effectiveness and superiority of the method.

Literature Study: Transfer Learning for COVID-19 Disease Analysis Based on Chest X-ray Dataset

The urgency of the impact of the COVID-19 disease that attacks people around the world encourages special research, especially in the field of artificial intelligence. This study aims to conduct a literature study related to the use of artificial intelligence, especially transfer learning in analyzing COVID-19 disease based on chest X-ray datasets. The research method of this research adapts the Preferred Reporting for Systematic Reviews and Meta-Analysis (PRISMA) guidelines. The results of the analysis of this data to answer research questions regarding the transfer learning model for the analysis of COVID-19 disease based on the chest X-ray dataset, it is known that the models used are MobileNet, Inception, VGG and ResNet. MobileNetV2 can be optimized by adding a global average pooling layer, dropout layer and dense layer and get an accuracy of 98.65%. InceptionV3 can be combined with Xception and get 98.8% accuracy. VGG-16 can be combined with ResNet-50 Xception and get 98.93% accuracy. ResNet-50 can be optimized by adding a dropout layer and a dense layer and getting an accuracy of 97.65%.

Comparative analysis of convolutional neural network and DenseNet121 transfer learning in agriculture focusing on crop leaf disease identification

PurposePlant cultivation holds a pivotal role in agriculture, necessitating precise disease identification for the overall health of plants. This research conducts a comprehensive comparative analysis between two prominent deep learning algorithms, convolutional neural network (CNN) and DenseNet121, with the goal of enhancing disease identification in tomato plant leaves.Design/methodology/approachThe dataset employed in this investigation is a fusion of primary data and publicly available data, covering 13 distinct disease labels and a total of 18,815 images for model training. The data pre-processing workflow prioritized activities such as normalizing pixel dimensions, implementing data augmentation and achieving dataset balance, which were subsequently followed by the modeling and testing phases.FindingsExperimental findings elucidated the superior performance of the DenseNet121 model over the CNN model in disease classification on tomato leaves. The DenseNet121 model attained a training accuracy of 98.27%, a validation accuracy of 87.47% and average recall, precision and F1-score metrics of 87, 88 and 87%, respectively. The ultimate aim was to implement the optimal classifier for a mobile application, namely Tanamin.id, and, therefore, DenseNet121 was the preferred choice.Originality/valueThe integration of private and public data significantly contributes to determining the optimal method. The CNN method achieves a training accuracy of 90.41% and a validation accuracy of 83.33%, whereas the DenseNet121 method excels with a training accuracy of 98.27% and a validation accuracy of 87.47%. The DenseNet121 architecture, comprising 121 layers, a global average pooling (GAP) layer and a dropout layer, showcases its effectiveness. Leveraging categorical_crossentropy as the loss function and utilizing the stochastic gradien descent (SGD) Optimizer with a learning rate of 0.001 guides the course of the training process. The experimental results unequivocally demonstrate the superior performance of DenseNet121 over CNN.

LMA-EEGNet: A Lightweight Multi-Attention Network for Neonatal Seizure Detection Using EEG signals

Neonatal epilepsy is an early postnatal brain disorder, and automatic seizure detection is crucial for timely diagnosis and treatment to reduce potential brain damage. This work proposes a novel Lightweight Multi-Attention Network, LMA-EEGNet, for diagnosing neonatal epileptic seizures from multi-channel EEG signals employing dilated depthwise separable convolution (DDS Conv) for feature extraction and using pointwise convolution followed by global average pooling for classification. The proposed approach substantially reduces the model size, number of parameters, and computational complexity, which are crucial for real-time detection and clinical diagnosis of neonatal epileptic seizures. LMA-EEGNet integrates temporal and spectral features through distinct temporal and spectral branches. The temporal branch uses DDS Conv to extract temporal features, enhanced by a channel attention mechanism. The spectral branch utilizes similar convolutions alongside a spatial attention mechanism to highlight key frequency components. Outputs from both branches are merged and processed through a pointwise convolution layer and a global average pooling layer for efficient neonatal seizure detection. Experimental results show that our model, with only 2471 parameters and a size of 23 KB, achieves an accuracy of 95.71% and an AUC of 0.9862, demonstrating its potential for practical deployment. This study provides an effective deep learning solution for the early detection of neonatal epileptic seizures, improving diagnostic accuracy and timeliness.

Transfer learning for acoustic cement bond evaluation: An image classification approach using acoustic variable Density log

Ensuring the integrity of wellbore cement is critical for the environmental protection, safety, and economic viability of oil, gas, geothermal, and Carbon Capture and Storage (CCS) operations. Acoustic logging is the most commonly used evaluation method, and interpreting its data requires significant expertise, often informs high-stakes decisions. Machine Learning (ML) has been used to help the data interpretation, but they are heavily reliant on meticulous feature engineering—a labor-intensive and expertise-driven task. Moreover, ML methods cannot tell which portions of input acoustic data are most important in the cement evaluation task. This level of analysis is vital, as it can significantly inform and improve future strategies for acoustic logging data collection and signal processing, ultimately improving the cement evaluation results. To address these challenges, we present a generalized workflow. First, we convert Variable Density Log (VDL) data into images using the Continuous Wavelet Transform (CWT). Then, we apply transfer learning for image classification to enhance the classification of wellbore cement isolation. In transfer learning stage, the images are processed using several pre-trained image classifiers—Xception, VGG16, MobileNetV2, and ResNet50—originally trained on the extensive ImageNet database containing over 14 million images. To adapt these models for our specific task, we added a global average pooling layer to reduce feature map dimensionality, followed by a fully connected layer with a Rectified Linear Unit (ReLU) activation function to introduce non-linearity and enhance learning capability. We replaced the original final layer of each classifier with a new three-neuron layer using softmax activation, tailored for multi-class classification. We preserved the general characteristics learned from the ImageNet dataset by freezing all original layers except for the newly added ones and compiled the models using the Adam optimizer. The performance of these adapted models was evaluated based on accuracy, loss, and F1 score. We applied this workflow to two distinct datasets: the first from a Norwegian wellbore and the second from a hydrocarbon well in China. In the Norwegian case, the Xception model achieved a classification accuracy of 97.4%, which was further refined to 99.0% through the implementation of Gradient-weighted Class Activation Mapping (Grad-CAM), revealing critical frequency ranges that traditional ML methods could not identify. In the Chinese case, employing the same models and workflow, the MobileNetV2 algorithm demonstrated exceptional performance, achieving a 99.6% accuracy in predicting cement isolation. In addition to achieving high accuracy, our results reveal that the VDL frequency content between 20 and 25 kHz is more critical for cement evaluation in the investigated wells than other frequency ranges, a level of analysis unattainable with traditional ML methods. Furthermore, our approach eliminates the need for deep knowledge in feature engineering, typically requisite in traditional ML, and operates effectively without any feature engineering. These results underscore the efficacy of our approach across diverse geological settings and highlight the potential of transfer learning to revolutionize cement evaluation automation, inspiring its broader application in the industry.

Explainability of three-dimensional convolutional neural networks for functional magnetic resonance imaging of Alzheimer's disease classification based on gradient-weighted class activation mapping.

Currently, numerous studies focus on employing fMRI-based deep neural networks to diagnose neurological disorders such as Alzheimer's Disease (AD), yet only a handful have provided results regarding explainability. We address this gap by applying several prevalent explainability methods such as gradient-weighted class activation mapping (Grad-CAM) to an fMRI-based 3D-VGG16 network for AD diagnosis to improve the model's explainability. The aim is to explore the specific Region of Interest (ROI) of brain the model primarily focuses on when making predictions, as well as whether there are differences in these ROIs between AD and normal controls (NCs). First, we utilized multiple resting-state functional activity maps including ALFF, fALFF, ReHo, and VMHC to reduce the complexity of fMRI data, which differed from many studies that utilized raw fMRI data. Compared to methods utilizing raw fMRI data, this manual feature extraction approach may potentially alleviate the model's burden. Subsequently, 3D-VGG16 were employed for AD classification, where the final fully connected layers were replaced with a Global Average Pooling (GAP) layer, aimed at mitigating overfitting while preserving spatial information within the feature maps. The model achieved a maximum of 96.4% accuracy on the test set. Finally, several 3D CAM methods were employed to interpret the models. In the explainability results of the models with relatively high accuracy, the highlighted ROIs were primarily located in the precuneus and the hippocampus for AD subjects, while the models focused on the entire brain for NC. This supports current research on ROIs involved in AD. We believe that explaining deep learning models would not only provide support for existing research on brain disorders, but also offer important referential recommendations for the study of currently unknown etiologies.

Classification of field wheat varieties based on a lightweight G-PPW-VGG11 model.

In agriculture, especially wheat cultivation, farmers often use multi-variety planting strategies to reduce monoculture-related harvest risks. However, the subtle morphological differences among wheat varieties make accurate discrimination technically challenging. Traditional variety classification methods, reliant on expert knowledge, are inefficient for modern intelligent agricultural management. Numerous existing classification models are computationally complex, memory-intensive, and difficult to deploy on mobile devices effectively. This study introduces G-PPW-VGG11, an innovative lightweight convolutional neural network model, to address these issues. G-PPW-VGG11 ingeniously combines partial convolution (PConv) and partially mixed depthwise separable convolution (PMConv), reducing computational complexity and feature redundancy. Simultaneously, incorporating ECANet, an efficient channel attention mechanism, enables precise leaf information capture and effective background noise suppression. Additionally, G-PPW-VGG11 replaces traditional VGG11's fully connected layers with two pointwise convolutional layers and a global average pooling layer, significantly reducing memory footprint and enhancing nonlinear expressiveness and training efficiency. Rigorous testing showed G-PPW-VGG11's superior performance, with an impressive 93.52% classification accuracy and only 1.79MB memory usage. Compared to VGG11, G-PPW-VGG11 showed a 5.89% increase in accuracy, 35.44% faster inference, and a 99.64% reduction in memory usage. G-PPW-VGG11 also surpasses traditional lightweight networks in classification accuracy and inference speed. Notably, G-PPW-VGG11 was successfully deployed on Android and its performance evaluated in real-world settings. The results showed an 84.67% classification accuracy with an average time of 291.04ms per image. This validates the model's feasibility for practical agricultural wheat variety classification, establishing a foundation for intelligent management. For future research, the trained model and complete dataset are made publicly available.

Impact location of metal structures based on time–frequency image features and deep residual network

Impact location detection plays an important role in the structural health monitoring of metal materials. However, the methods of metal material impact location detection based on physical analysis are often limited by the extraction accuracy of some parameters such as material and structure parameters and time difference calculation. Therefore, this paper develops a deep residual network method for impact location detection, time–frequency characteristic deep residual network (TF-DRN). This method takes the four-channel short-time Fourier transform time–frequency graph as input, uses the unique residual network architecture to automatically extract the advanced features, and then uses the global average pooling layer and the full connection layer to establish the mapping between the advanced features and the impact location, so as to detect the impact location. By introducing regularization and batch normalization, the problems of gradient disappearance and gradient explosion are alleviated, and the generalization and efficiency of impact location detection are further improved. The experimental results show that on an 800 mm × 800 mm × 2.5 mm aluminum plate, the average error of the validation set and the test set are 0.85 cm and 1.33 cm respectively, and the performance of the method is significantly better than that of CNN, ResNet18 and ResNet33 networks.

MACN: A cascade defect detection for complex background based on mixture attention mechanism

AbstractDefect detection in complex background is a critical issue. To address this issue, this paper proposes the mixture attention mechanism cascade network, in which the new channel attention network is linked with the spatial attention network to create an effective mixed attention network that takes advantage of their respective advantages, adaptively suppresses background noise, and highlights defect features. To ensure the efficiency and effectiveness of effective mixed attention network, the new channel attention network splices the output features of the global average pooling layer and the global maximum pooling layer and then sends the spliced features into a shared network, which is a one‐dimensional convolutional network, and uses cross‐channel interaction for fusion. Furthermore, in order to provide more discriminative feature representation, the authors extract the intermediate features of the region proposal network and input them into effective mixed attention network. Finally, the cascade head is used to refine the predicted bounding box to achieve high‐quality defect location. To demonstrate the superiority and usefulness of this method, it is compared to the latest method using widely used PCB and NEU data sets. A large number of trials demonstrate that this strategy outperforms other methods for detecting defects in complicated backgrounds.

INDIAN SIGN LANGUAGE RECOGNITION USING CNN WITH SPATIAL PYRAMID AND GLOBAL AVERAGE POOLING

Sign language recognition plays a pivotal role in bridging communication gaps for the speech and hearing-impaired community. In this research, we present an innovative approach to enhance the accuracy and effectiveness of Indian sign language recognition through the integration of convolutional neural networks with spatial pyramid and global average pooling layers. The objective of this study is to address the intricate challenges by exploiting multi-scale feature representations and global context information. This architecture incorporates spatial pyramid layer to capture fine-grained spatial information across multiple scales. Additionally, global average pooling layer is introduced to consolidate context-aware features, further improving the model's discriminative power. Experiments are conducted using a standard dataset, encompassing a diverse range of gestures. The results demonstrate a significant improvement in recognition accuracy compared to conventional models. Incorporating spatial pyramid and global average pooling layers enables our model to effectively recognize the gestures, even in challenging scenarios with variations in lighting, complex background, and signer-dependent factors by achieving high accuracy. This research not only contributes to the advancement of ISL recognition technology but also holds promise for practical applications in real-world communication and assistive technologies.

Music-evoked emotions classification using vision transformer in EEG signals.

The field of electroencephalogram (EEG)-based emotion identification has received significant attention and has been widely utilized in both human-computer interaction and therapeutic settings. The process of manually analyzing electroencephalogram signals is characterized by a significant investment of time and work. While machine learning methods have shown promising results in classifying emotions based on EEG data, the task of extracting distinct characteristics from these signals still poses a considerable difficulty. In this study, we provide a unique deep learning model that incorporates an attention mechanism to effectively extract spatial and temporal information from emotion EEG recordings. The purpose of this model is to address the existing gap in the field. The implementation of emotion EEG classification involves the utilization of a global average pooling layer and a fully linked layer, which are employed to leverage the discernible characteristics. In order to assess the effectiveness of the suggested methodology, we initially gathered a dataset of EEG recordings related to music-induced emotions. Subsequently, we ran comparative tests between the state-of-the-art algorithms and the method given in this study, utilizing this proprietary dataset. Furthermore, a publicly accessible dataset was included in the subsequent comparative trials. The experimental findings provide evidence that the suggested methodology outperforms existing approaches in the categorization of emotion EEG signals, both in binary (positive and negative) and ternary (positive, negative, and neutral) scenarios.