EEG-based emotion recognition using graph convolutional neural network with dual attention mechanism.

EEG-based emotion recognition is becoming crucial in brain-computer interfaces (BCI). Currently, most researches focus on improving accuracy, while neglecting further research on the interpretability of models, we are committed to analyzing the impact of different brain regions and signal frequency bands on emotion generation based on graph structure. Therefore, this paper proposes a method named Dual Attention Mechanism Graph Convolutional Neural Network (DAMGCN). Specifically, we utilize graph convolutional neural networks to model the brain network as a graph to extract representative spatial features. Furthermore, we employ the self-attention mechanism of the Transformer model which allocates more electrode channel weights and signal frequency band weights to important brain regions and frequency bands. The visualization of attention mechanism clearly demonstrates the weight allocation learned by DAMGCN. During the performance evaluation of our model on the DEAP, SEED, and SEED-IV datasets, we achieved the best results on the SEED dataset, showing subject-dependent experiments' accuracy of 99.42% and subject-independent experiments' accuracy of 73.21%. The results are demonstrably superior to the accuracies of most existing models in the realm of EEG-based emotion recognition.

DEAF-Net: Detail-Enhanced Attention Feature Fusion Network for Retinal Vessel Segmentation.

Retinal vessel segmentation is crucial for the diagnosis of ophthalmic and cardiovascular diseases. However, retinal vessels are densely and irregularly distributed, with many capillaries blending into the background, and exhibit low contrast. Moreover, the encoder-decoder-based network for retinal vessel segmentation suffers from irreversible loss of detailed features due to multiple encoding and decoding, leading to incorrect segmentation of the vessels. Meanwhile, the single-dimensional attention mechanisms possess limitations, neglecting the importance of multidimensional features. To solve these issues, in this paper, we propose a detail-enhanced attention feature fusion network (DEAF-Net) for retinal vessel segmentation. First, the detail-enhanced residual block (DERB) module is proposed to strengthen the capacity for detailed representation, ensuring that intricate features are efficiently maintained during the segmentation of delicate vessels. Second, the multidimensional collaborative attention encoder (MCAE) module is proposed to optimize the extraction of multidimensional information. Then, the dynamic decoder (DYD) module is introduced to preserve spatial information during the decoding process and reduce the information loss caused by upsampling operations. Finally, the proposed detail-enhanced feature fusion (DEFF) module composed of DERB, MCAE and DYD modules fuses feature maps from both encoding and decoding and achieves effective aggregation of multi-scale contextual information. The experiments conducted on the datasets of DRIVE, CHASEDB1, and STARE, achieving Sen of 0.8305, 0.8784, and 0.8654, and AUC of 0.9886, 0.9913, and 0.9911 on DRIVE, CHASEDB1, and STARE, respectively, demonstrate the performance of our proposed network, particularly in the segmentation of fine retinal vessels.

A novel graph neural network method for Alzheimer's disease classification.

Alzheimer's disease (AD) is a chronic neurodegenerative disease. Early diagnosis are very important to timely treatment and delay the progression of the disease. In the past decade, many computer-aided diagnostic (CAD) algorithms have been proposed for classification of AD. In this paper, we propose a novel graph neural network method, termed Brain Graph Attention Network (BGAN) for classification of AD. First, brain graph data are used to model classification of AD as a graph classification task. Second, a local attention layer is designed to capture and aggregate messages of interactions between node neighbors. And, a global attention layer is introduced to obtain the contribution of each node for graph representation. Finally, using the BGAN to implement AD classification. We train and test on two open public databases for AD classification task. Compared to classic models, the experimental results show that our model is superior to six classic models. We demonstrate that BGAN is a powerful classification model for AD. In addition, our model can provide an analysis of brain regions in order to judge which regions are related to AD disease and which regions are related to AD progression.

Memristive patch attention neural network for facial expression recognition and edge computing.

Facial expression recognition has made a significant progress as a result of the advent of more and more convolutional neural networks (CNN). However, with the improvement of CNN, the models continues to get deeper and larger so as to a greater focus on the high-level features of the image and the low-level features tend to be lost. Because of the reason above, the dependence of low-level features between different areas of the face often cannot be summarized. In response to this problem, we propose a novel network based on the CNN model. To extract long-range dependencies of low-level features, multiple attention mechanisms has been introduced into the network. In this paper, the patch attention mechanism is designed to obtain the dependence between low-level features of facial expressions firstly. After fusion, the feature maps are input to the backbone network incorporating convolutional block attention module (CBAM) to enhance the feature extraction ability and improve the accuracy of facial expression recognition, and achieve competitive results on three datasets CK+ (98.10%), JAFFE (95.12%) and FER2013 (73.50%). Further, according to the PA Net designed in this paper, a hardware friendly implementation scheme is designed based on memristor crossbars, which is expected to provide a software and hardware co-design scheme for edge computing of personal and wearable electronic products.

M 3: using mask-attention and multi-scale for multi-modal brain MRI classification.

Introduction: Brain diseases, particularly the classification of gliomas and brain metastases and the prediction of HT in strokes, pose significant challenges in healthcare. Existing methods, relying predominantly on clinical data or imaging-based techniques such as radiomics, often fall short in achieving satisfactory classification accuracy. These methods fail to adequately capture the nuanced features crucial for accurate diagnosis, often hindered by noise and the inability to integrate information across various scales. Methods: We propose a novel approach that mask attention mechanisms with multi-scale feature fusion for Multimodal brain disease classification tasks, termed M 3, which aims to extract features highly relevant to the disease. The extracted features are then dimensionally reduced using Principal Component Analysis (PCA), followed by classification with a Support Vector Machine (SVM) to obtain the predictive results. Results: Our methodology underwent rigorous testing on multi-parametric MRI datasets for both brain tumors and strokes. The results demonstrate a significant improvement in addressing critical clinical challenges, including the classification of gliomas, brain metastases, and the prediction of hemorrhagic stroke transformations. Ablation studies further validate the effectiveness of our attention mechanism and feature fusion modules. Discussion: These findings underscore the potential of our approach to meet and exceed current clinical diagnostic demands, offering promising prospects for enhancing healthcare outcomes in the diagnosis and treatment of brain diseases.

Multi-disease X-ray Image Classification of the Chest Based on Global and Local Fusion Adaptive Networks.

BACKGROUND: Chest X-ray image classification for multiple diseases is an important research direction in the field of computer vision and medical image processing. It aims to utilize advanced image processing techniques and deep learning algorithms to automatically analyze and identify X-ray images, determining whether specific pathologies or structural abnormalities exist in the images. OBJECTIVE: We present the MMPDenseNet network designed specifically for chest multi-label disease classification. METHODS: Initially, the network employs the adaptive activation function Meta-ACON to enhance feature representation. Subsequently, the network incorporates a multi-head self-attention mechanism, merging the conventional convolutional neural network with the Transformer, thereby bolstering the ability to extract both local and global features. Ultimately, the network integrates a pyramid squeeze attention module to capture spatial information and enrich the feature space. RESULTS: The concluding experiment yielded an average AUC of 0.898, marking an average accuracy improvement of 0.6% over the baseline model. When compared with the original network, the experimental results highlight that MMPDenseNet considerably elevates the classification accuracy of various chest diseases. CONCLUSION: It can be concluded that the network, thus, holds substantial value for clinical applications.

Cross-modal knowledge distillation for continuous sign language recognition.

Continuous Sign Language Recognition (CSLR) is a task which converts a sign language video into a gloss sequence. The existing deep learning based sign language recognition methods usually rely on large-scale training data and rich supervised information. However, current sign language datasets are limited, and they are only annotated at sentence-level rather than frame-level. Inadequate supervision of sign language data poses a serious challenge for sign language recognition, which may result in insufficient training of sign language recognition models. To address above problems, we propose a cross-modal knowledge distillation method for continuous sign language recognition, which contains two teacher models and one student model. One of the teacher models is the Sign2Text dialogue teacher model, which takes a sign language video and a dialogue sentence as input and outputs the sign language recognition result. The other teacher model is the Text2Gloss translation teacher model, which targets to translate a text sentence into a gloss sequence. Both teacher models can provide information-rich soft labels to assist the training of the student model, which is a general sign language recognition model. We conduct extensive experiments on multiple commonly used sign language datasets, i.e., PHOENIX 2014T, CSL-Daily and QSL, the results show that the proposed cross-modal knowledge distillation method can effectively improve the sign language recognition accuracy by transferring multi-modal information from teacher models to the student model. Code is available at https://github.com/glq-1992/cross-modal-knowledge-distillation_new.

RS-Dseg: semantic segmentation of high-resolution remote sensing images based on a diffusion model component with unsupervised pretraining.

Semantic segmentation plays a crucial role in interpreting remote sensing images, especially in high-resolution scenarios where finer object details, complex spatial information and texture structures exist. To address the challenge of better extracting semantic information and ad-dressing class imbalance in multiclass segmentation, we propose utilizing diffusion models for remote sensing image semantic segmentation, along with a lightweight classification module based on a spatial-channel attention mechanism. Our approach incorporates unsupervised pretrained components with a classification module to accelerate model convergence. The diffusion model component, built on the UNet architecture, effectively captures multiscale features with rich contextual and edge information from images. The lightweight classification module, which leverages spatial-channel attention, focuses more efficiently on spatial-channel regions with significant feature information. We evaluated our approach using three publicly available datasets: Postdam, GID, and Five Billion Pixels. In the test of three datasets, our method achieved the best results. On the GID dataset, the overall accuracy was 96.99%, the mean IoU was 92.17%, and the mean F1 score was 95.83%. In the training phase, our model achieved good performance after only 30 training cycles. Compared with other models, our method reduces the number of parameters, improves the training speed, and has obvious performance advantages.

Lightweight safflower cluster detection based on YOLOv5.

The effective detection of safflower in the field is crucial for implementing automated visual navigation and harvesting systems. Due to the small physical size of safflower clusters, their dense spatial distribution, and the complexity of field scenes, current target detection technologies face several challenges in safflower detection, such as insufficient accuracy and high computational demands. Therefore, this paper introduces an improved safflower target detection model based on YOLOv5, termed Safflower-YOLO (SF-YOLO). This model employs Ghost_conv to replace traditional convolution blocks in the backbone network, significantly enhancing computational efficiency. Furthermore, the CBAM attention mechanism is integrated into the backbone network, and a combined L C I O U + N W D loss function is introduced to allow for more precise feature extraction, enhanced adaptive fusion capabilities, and accelerated loss convergence. Anchor boxes, updated through K-means clustering, are used to replace the original anchors, enabling the model to better adapt to the multi-scale information of safflowers in the field. Data augmentation techniques such as Gaussian blur, noise addition, sharpening, and channel shuffling are applied to the dataset to maintain robustness against variations in lighting, noise, and visual angles. Experimental results demonstrate that SF-YOLO surpasses the original YOLOv5s model, with reductions in GFlops and Params from 15.8 to 13.2 G and 7.013 to 5.34 M, respectively, representing decreases of 16.6% and 23.9%. Concurrently, SF-YOLO's mAP0.5 increases by 1.3%, reaching 95.3%. This work enhances the accuracy of safflower detection in complex agricultural environments, providing a reference for subsequent autonomous visual navigation and automated non-destructive harvesting technologies in safflower operations.

DAC-Net: A light-weight U-shaped network based efficient convolution and attention for thyroid nodule segmentation.

Recently, there has been a focused effort to improve the efficiency of thyroid nodule segmentation algorithms. This endeavor has resulted in the development of increasingly complex modules, such as the Transformer, leading to models with a higher number of parameters and computing requirements. Sophisticated models have difficulties in being implemented in clinical medicine platforms because of limited resources. DAC-Net is a Lightweight U-shaped network created to achieve high performance in segmenting thyroid nodules. Our method consists of three main components: DWSE, which combines depthwise convolution and squeeze-excitation block to enhance feature extraction and connections between samples; ADA, which includes Split Atrous and Dual Attention to extract global and local feature information from various viewpoints; and CSSC, which involves channel- scale and spatial-scale connections. This module enables the fusing of multi-stage features at global and local levels, producing feature maps at different channel and geographical scales, delivering a streamlined integration of multi-scale information. Combining these three components in our U- shaped design allows us to achieve competitive performance while also decreasing the number of parameters and computing complexity. Several experiments were conducted on the DDTI and TN3K datasets. The experimental results demonstrate that our model outperforms state-of-the-art thyroid nodule segmentation architectures in terms of segmentation performance. Our model not only reduces the number of parameters and computing expenses by 73x and 56x, respectively, but also exceeds TransUNet in segmentation performance. The source code is accessible at https://github.com/Phil-y/DAC-Net.

Global remote feature modulation end-to-end detection.

Object detector based on fully convolutional network achieves excellent performance. However, existing detection algorithms still face challenges such as low detection accuracy in dense scenes and issues with occlusion of dense targets. To address these two challenges, we propose an Global Remote Feature Modulation End-to-End (GRFME2E) detection algorithm. In the feature extraction phase of our algorithm, we introduces the Concentric Attention Feature Pyramid Network (CAFPN). The CAFPN captures direction-aware and position-sensitive information, as well as global remote dependencies of features in deep layers by combining Coordinate Attention and Multilayer Perceptron. These features are used to modulate the front-end shallow features, enhancing inter-layer feature adjustment to obtain comprehensive and distinctive feature representations.In the detector part, we introduce the Two-Stage Detection Head (TS Head). This head employs the First-One-to-Few (F-O2F) module to detect slightly or unobstructed objects. Additionally, it uses masks to suppress already detected instances, and then feeds them to the Second-One-to-Few (S-O2F) module to identify those that are heavily occluded. The results from both detection stages are merged to produce the final output, ensuring the detection of objects whether they are slightly obscured, unobstructed, or heavily occluded. Experimental results on the pig detection dataset demonstrate that our GRFME2E achieves an accuracy of 98.4%. In addition, more extensive experimental results show that on the CrowdHuman dataset, our GRFME2E achieves 91.8% and outperforms other methods.

Time series prediction model using LSTM-Transformer neural network for mine water inflow.

Mine flooding accidents have occurred frequently in recent years, and the predicting of mine water inflow is one of the most crucial flood warning indicators. Further, the mine water inflow is characterized by non-linearity and instability, making it difficult to predict. Accordingly, we propose a time series prediction model based on the fusion of the Transformer algorithm, which relies on self-attention, and the LSTM algorithm, which captures long-term dependencies. In this paper, Baotailong mine water inflow in Heilongjiang Province is used as sample data, and the sample data is divided into different ratios of the training set and test set in order to obtain optimal prediction results. In this study, we demonstrate that the LSTM-Transformer model exhibits the highest training accuracy when the ratio is 7:3. To improve the efficiency of search, the combination of random search and Bayesian optimization is used to determine the network model parameters and regularization parameters. Finally, in order to verify the accuracy of the LSTM-Transformer model, the LSTM-Transformer model is compared with LSTM, CNN, Transformer and CNN-LSTM models. The results prove that LSTM-Transformer has the highest prediction accuracy, and all the indicators of its model are well improved.

Microwave detection technique combined with deep learning algorithm facilitates quantitative analysis of heavy metal Pb residues in edible oils.

The heavy metal content in edible oils is intricately associated with their suitability for human consumption. In this study, standard soybean oil was used as a sample to quantify the specified concentration of heavy metals using microwave sensing technique. In addition, an attention-based deep residual neural network model was developed as an alternative to traditional modeling methods for predicting heavy metals in edible oils. In the process of microwave data processing, this work continued to discuss the impact of depth on convolutional neural networks. The results demonstrated that the proposed attention-based residual network model outperforms all other deep learning models in all metrics. The performance of this model was characterized by a coefficient of determination (R2) of 0.9605, a relative prediction deviation (RPD) of 5.0479, and a root mean square error (RMSE) of 3.1654 mg/kg. The research findings indicate that the combination of microwave detection technology and chemometrics holds significant potential for assessing heavy metal levels in edible oils.

RSANMDA: Resampling based subview attention network for miRNA-disease association prediction.

Many studies have demonstrated the importance of accurately identifying miRNA-disease associations (MDAs) for understanding disease mechanisms. However, the number of known MDAs is significantly fewer than the unknown pairs. Here, we propose RSANMDA, a subview attention network for predicting MDAs. We first extract miRNA and disease features from multiple similarity matrices. Next, using resampling techniques, we generate different subviews from known MDAs. Each subview undergoes multi-head graph attention to capture its features, followed by semantic attention to integrate features across subviews. Finally, combining raw and training features, we use a multilayer scoring perceptron for prediction. In the experimental section, we conducted comparative experiments with other advanced models on both HMDD v2.0 and HMDD v3.2 datasets. We also performed a series of ablation studies and parameter tuning exercises. Comprehensive experiments conclusively demonstrate the superiority of our model. Case studies on lung, breast, and esophageal cancers further validate our method's predictive capability for identifying disease-related miRNAs.

DLCNBC-SA: a model for assessing axillary lymph node metastasis status in early breast cancer patients.

Background: Axillary lymph node (ALN) status is a crucial prognostic indicator for breast cancer metastasis, with manual interpretation of whole slide images (WSIs) being the current standard practice. However, this method is subjective and time-consuming. Recent advancements in deep learning-based methods for medical image analysis have shown promise in improving clinical diagnosis. This study aims to leverage these technological advancements to develop a deep learning model based on features extracted from primary tumor biopsies for preoperatively identifying ALN metastasis in early-stage breast cancer patients with negative nodes. Methods: We present DLCNBC-SA, a deep learning-based network specifically tailored for core needle biopsy and clinical data feature extraction, which integrates a self-attention mechanism (CNBC-SA). The proposed model consists of a feature extractor based on convolutional neural network (CNN) and an improved self-attention mechanism module, which can preserve the independence of features in WSIs for analysis and enhancement to provide rich feature representation. To validate the performance of the proposed model, we conducted comparative experiments and ablation studies using publicly available datasets, and verification was performed through quantitative analysis. Results: The comparative experiment illustrates the superior performance of the proposed model in the task of binary classification of ALNs, as compared to alternative methods. Our method achieved outstanding performance [area under the curve (AUC): 0.882] in this task, significantly surpassing the state-of-the-art (SOTA) method on the same dataset (AUC: 0.862). The ablation experiment reveals that incorporating RandomRotation data augmentation technology and utilizing Adadelta optimizer can effectively enhance the performance of the proposed model. Conclusions: The experimental results demonstrate that the model proposed in this paper outperforms the SOTA model on the same dataset, thereby establishing its reliability as an assistant for pathologists in analyzing WSIs of breast cancer. Consequently, it significantly enhances both the efficiency and accuracy of doctors during the diagnostic process.

A mutual reconstruction network model for few-shot classification of histological images: addressing interclass similarity and intraclass diversity.

Background: The automated classification of histological images is crucial for the diagnosis of cancer. The limited availability of well-annotated datasets, especially for rare cancers, poses a significant challenge for deep learning methods due to the small number of relevant images. This has led to the development of few-shot learning approaches, which bear considerable clinical importance, as they are designed to overcome the challenges of data scarcity in deep learning for histological image classification. Traditional methods often ignore the challenges of intraclass diversity and interclass similarities in histological images. To address this, we propose a novel mutual reconstruction network model, aimed at meeting these challenges and improving the few-shot classification performance of histological images. Methods: The key to our approach is the extraction of subtle and discriminative features. We introduce a feature enhancement module (FEM) and a mutual reconstruction module to increase differences between classes while reducing variance within classes. First, we extract features of support and query images using a feature extractor. These features are then processed by the FEM, which uses a self-attention mechanism for self-reconstruction of features, enhancing the learning of detailed features. These enhanced features are then input into the mutual reconstruction module. This module uses enhanced support features to reconstruct enhanced query features and vice versa. The classification of query samples is based on weighted calculations of the distances between query features and reconstructed query features and between support features and reconstructed support features. Results: We extensively evaluated our model using a specially created few-shot histological image dataset. The results showed that in a 5-way 10-shot setup, our model achieved an impressive accuracy of 92.09%. This is a 23.59% improvement in accuracy compared to the model-agnostic meta-learning (MAML) method, which does not focus on fine-grained attributes. In the more challenging, 5-way 1-shot setting, our model also performed well, demonstrating a 18.52% improvement over the ProtoNet, which does not address this challenge. Additional ablation studies indicated the effectiveness and complementary nature of each module and confirmed our method's ability to parse small differences between classes and large variations within classes in histological images. These findings strongly support the superiority of our proposed method in the few-shot classification of histological images. Conclusions: The mutual reconstruction network provides outstanding performance in the few-shot classification of histological images, successfully overcoming the challenges of similarities between classes and diversity within classes. This marks a significant advancement in the automated classification of histological images.

Degradation prediction of fuel cell systems based on different operating conditions in dynamic cycling condition.

In this paper, the degradation of PEMFC under different operating conditions in dynamic cycle condition is studied. Firstly, according to the failure mechanism of PEMFC, various operating conditions in dynamic cycle condition are classified, and the health indexes are established. Simultaneously, the rates and degrees of the output voltage decline of the PEMFC under different operating conditions during the dynamic cycling process were compared. Then, a model based on variational mode decomposition and long short-term memory with attention mechanism (VMD-LSTM-ATT) is proposed. Aiming at the performance of PEMFC is affected by the external operation, VMD is used to capture the global information and details, and filter out interference information. To improve the prediction accuracy, ATT is used to assign weight to the features. Finally, in order to verify the effectiveness and superiority of VMD-LSTM-ATT, we respectively apply it to three current conditions under dynamic cycle conditions. The experimental results show that under the same test conditions, RMSE of VMD-LSTM-ATT is increased by 56.11 % and MAE is increased by 28.26 % compared with GRU attention. Compared with SVM, RNN, LSTM and LSTM-ATT, RMSE of VMD-LSTM-ATT is at least 17.26 % higher and MAE is at least 5.65 % higher.

Maize plant height automatic reading of measurement scale based on improved YOLOv5 lightweight model.

Background: Plant height is a significant indicator of maize phenotypic morphology, and is closely related to crop growth, biomass, and lodging resistance. Obtaining the maize plant height accurately is of great significance for cultivating high-yielding maize varieties. Traditional measurement methods are labor-intensive and not conducive to data recording and storage. Therefore, it is very essential to implement the automated reading of maize plant height from measurement scales using object detection algorithms. Method: This study proposed a lightweight detection model based on the improved YOLOv5. The MobileNetv3 network replaced the YOLOv5 backbone network, and the Normalization-based Attention Module attention mechanism module was introduced into the neck network. The CioU loss function was replaced with the EioU loss function. Finally, a combined algorithm was used to achieve the automatic reading of maize plant height from measurement scales. Results: The improved model achieved an average precision of 98.6%, a computational complexity of 1.2 GFLOPs, and occupied 1.8 MB of memory. The detection frame rate on the computer was 54.1 fps. Through comparisons with models such as YOLOv5s, YOLOv7 and YOLOv8s, it was evident that the comprehensive performance of the improved model in this study was superior. Finally, a comparison between the algorithm's 160 plant height data obtained from the test set and manual readings demonstrated that the relative error between the algorithm's results and manual readings was within 0.2 cm, meeting the requirements of automatic reading of maize height measuring scale.

A Shape-Consistent Deep-Learning Segmentation Architecture for Low-Quality and High-Interference Myocardial Contrast Echocardiography.

OBJECTIVE: Myocardial contrast echocardiography (MCE) plays a crucial role in diagnosing ischemia, infarction, masses and other cardiac conditions. In the realm of MCE image analysis, accurate and consistent myocardial segmentation results are essential for enabling automated analysis of various heart diseases. However, current manual diagnostic methods in MCE suffer from poor repeatability and limited clinical applicability. MCE images often exhibit low quality and high noise due to the instability of ultrasound signals, while interference structures can further disrupt segmentation consistency. METHODS: To overcome these challenges, we proposed a deep-learning network for the segmentation of MCE. This architecture leverages dilated convolutions to capture high-scale information without sacrificing positional accuracy and modifies multi-head self-attention to enhance global context and ensure consistency, effectively overcoming issues related to low image quality and interference. Furthermore, we also adapted the cascade application of transformers with convolutional neural networks for improved segmentation in MCE. RESULTS: In our experiments, our architecture achieved the best Dice score of 84.35% for standard MCE views compared with that of several state-of-the-art segmentation models. For non-standard views and frames with interfering structures (mass), our models also attained the best Dice scores of 83.33% and 83.97%, respectively. CONCLUSION: These studies proved that our architecture is of excellent shape consistency and robustness, which allows it to deal with segmentation of various types of MCE. Our relatively precise and consistent myocardial segmentation results provide fundamental conditions for the automated analysis of various heart diseases, with the potential to discover underlying pathological features and reduce healthcare costs.

Protein-peptide binding residue prediction based on protein language models and cross-attention mechanism.

Accurate identifications of protein-peptide binding residues are essential for protein-peptide interactions and advancing drug discovery. To address this problem, extensive research efforts have been made to design more discriminative feature representations. However, extracting these explicit features usually depend on third-party tools, resulting in low computational efficacy and suffering from low predictive performance. In this study, we design an end-to-end deep learning-based method, E2EPep, for protein-peptide binding residue prediction using protein sequence only. E2EPep first employs and fine-tunes two state-of-the-art pre-trained protein language models that can extract two different high-latent feature representations from protein sequences relevant for protein structures and functions. A novel feature fusion module is then designed in E2EPep to fuse and optimize the above two feature representations of binding residues. In addition, we have also design E2EPep+, which integrates E2EPep and PepBCL models, to improve the prediction performance. Experimental results on two independent testing data sets demonstrate that E2EPep and E2EPep + could achieve the average AUC values of 0.846 and 0.842 while achieving an average Matthew's correlation coefficient value that is significantly higher than that of existing most of sequence-based methods and comparable to that of the state-of-the-art structure-based predictors. Detailed data analysis shows that the primary strength of E2EPep lies in the effectiveness of feature representation using cross-attention mechanism to fuse the embeddings generated by two fine-tuned protein language models. The standalone package of E2EPep and E2EPep + can be obtained at https://github.com/ckx259/E2EPep.git for academic use only.