Electrocardiogram (ECG) is an efficient and simple method for the diagnosis of cardiovascular diseases and has been widely used in clinical practice. Because of the shortage of professional cardiologists and the popularity of electrocardiograms, accurate and efficient arrhythmia detection has become a hot research topic. In this paper, we propose a new multi-task deep neural network, which includes a shared low-level feature extraction module (i.e., SE-ResNet) and a task-specific classification module. Contextual Transformer (CoT) block is introduced in the classification module to dynamically model the local and global information of ECG feature sequence. The proposed method was evaluated on public CPSC2018 and PTB-XL datasets and achieved an average F1 score of 0.827 on the CPSC2018 dataset and an average F1 score of 0.833 on the PTB-XL dataset.

Due to insufficient samples in rare arrhythmia classes, current research in ECG classification has reported poor performance in rare arrhythmia classes and difficulties in promptly generalizing to novel, unseen classes. This paper introduces a novel multi-level metric fusion framework specifically designed for few-shot electrocardiogram (ECG) classification. The framework aims to extract multi-level deep local descriptors and learn the intricate relationships between samples across different levels, thereby enhancing the few-shot ECG classification capability. In the proposed method, the episodic training mechanism based on N-way K-shot task is introduced to effectively train deep neural networks with limited samples. Then a unique multi-level metric fusion framework is developed for similarity learning, which jointly utilizes heartbeat-level and segment-level features for metric learning. Furthermore, the multi-level metric fusion module integrates weighted fusion of heartbeat-level and segment-level scores along the branch dimension, forming a a global decision similarity.The proposed method was evaluated on a few-shot ECG database, demonstrating remarkable performance with average accuracies of 72.74%, 69.81%, and 49.84% for 5-way 10-shot, 5-shot, and 1-shot scenarios, respectively. These results significantly surpass existing methods, highlighting the effectiveness of the proposed approach. The study presents a straightforward yet highly effective solution for multi-lead ECG classification, particularly suitable for few-shot scenarios. Furthermore, the proposed multi-level metric fusion module effectively leverages multi-level features, demonstrating its potential for improved generalization and accuracy in ECG classification tasks. This is the first study to achieve wide-range class classification in few-shot scenarios without relying on auxiliary techniques such as transfer learning or pre-training, and demonstrating important impact to clinical applications.

Researchers have introduced side information such as social networks or knowledge graphs to alleviate the problems of data sparsity and cold starts in recommendation systems. However, most of the methods ignore the exploration of feature differentiation aspects in the knowledge propagation process. To solve the above problem, we propose a new attention recommendation method based on an enhanced knowledge propagation perception. Specifically, to capture user preferences in a fine-grained manner in a knowledge graph, an asymmetric semantic attention mechanism is adopted. It identifies the influence of propagation neighbors on user preferences through a more precise representation of the preference semantics for head and tail entities. Furthermore, in consideration of the memory and generalization of different propagation depth features and adaptively adjusting the propagation weights, a new propagation feature exploration framework is designed. The performance of the proposed model is validated by two real-world datasets. The baseline model averagely increases by 9.65% and 9.15% for the Area Under Curve (AUC) and Accuracy (ACC) indicators, which proves the effectiveness of the model.

Crowd counting using RGB images has made significant progress in practical applications, but it often struggles in low-light conditions. Recent advancements in infrared sensor technology have introduced thermal imaging as a powerful complement to RGB-based methods, allowing for more accurate identification of individuals in dark environments. However, effectively integrating RGB and thermal data remains challenging due to their inherent differences, and both image types are susceptible to background noise, which can blur distinctions between pedestrians and background elements. To address these challenges, we propose a novel three-stream RGB-T crowd counting network, termed the Feature Transfer and Enhanced Attention network (FENet). It incorporates two key modules: the Feature-Integrated Transfer (FIT) module and the Attention-Enhanced (AE) module. The FIT module is designed to fuse modality-specific features from both the RGB and thermal streams into a unified representation, facilitating progressively refined cross-modal complementary information through continuous feature propagation. To increase the sensitivity to crowd location, the AE module is introduced to explore the geometric dependencies between the head and surrounding regions (such as background noise) by perceiving the influence of different regions on the head, ensuring that the head receives a higher attention weight. Extensive experiments on the RGBT-CC dataset demonstrate that our method achieves superior performance in RGB-T crowd counting. Moreover, our approach also shows excellent performance on the ShanghaiTechRGBD dataset, confirming its generalizability across different modalities.