Sign language recognition (SLR) presents significant challenges due to the complexities in modeling the intricate interactions between hand gestures and other body movements, as well as the presence of redundant and noisy temporal features. Existing Graph Convolutional Network (GCN)-based methods, while promising in capturing skeletal dynamics, are limited by their reliance on fixed skeletal structures and uniform frame importance, which hinder their effectiveness in addressing these issues. To overcome these challenges, we propose a novel Hand-Aware Masked Graph Convolutional Network (HAM-GCN), which introduces two key innovations: an adaptive dynamic hand graph and an adaptive masking mechanism. The hand graph effectively captures fine-grained interactions between hand gestures and body movements, enhancing the accuracy of sign language representations. The adaptive masking mechanism, driven by hand-specific confidence scores, dynamically adjusts the importance of individual frames, emphasizing informative ones while suppressing noisy or redundant features. We validate HAM-GCN through extensive experiments on three publicly available sign language datasets, achieving state-of-the-art performance and surpassing existing methods. The source code is publicly available at https://github.com/Gavin113/HAM-GCN.

For the skeleton-based gesture recognition, graph convolutional networks (GCNs) have achieved remarkable performance since the human skeleton is a natural graph. However, the biological structure might not be the crucial one for motion analysis. Also, spatial differential information like joint distance and angle between bones may be overlooked during the graph convolution. In this article, we focus on obtaining meaningful joint groups and extracting their discriminative features by the path signature (PS) theory. Firstly, to characterize the constraints and dependencies of various joints, we propose three types of paths, i.e., spatial, temporal, and learnable path. Especially, a learnable path generation mechanism can group joints together that are not directly connected or far away, according to their kinematic characteristic. Secondly, to obtain informative and compact features, a deep integration of PS with few parameters are introduced. All the computational process is packed into two modules, i.e., spatial-temporal path signature module (ST-PSM) and learnable path signature module (L-PSM) for the convenience of utilization. They are plug-and-play modules available for any neural network like CNNs and GCNs to enhance the feature extraction ability. Extensive experiments have conducted on three mainstream datasets (ChaLearn 2013, ChaLearn 2016, and AUTSL). We achieved the state-of-the-art results with simpler framework and much smaller model size. By inserting our two modules into the several GCN-based networks, we can observe clear improvements demonstrating the great effectiveness of our proposed method.

With the increasing prevalence of cardiovascular diseases, accurate prediction and monitoring of human blood pressure have become increasingly important. In the context of continuous interdisciplinary advancements, non-invasive blood pressure monitoring techniques have also seen rapid development. However, most existing research relies on conventional computational or machine learning algorithms for feature engineering and is validated only on a small number of test sets, resulting in limited generalization performance and a lack of universality. In this study, we propose an innovative approach that combines multimodal data generation and feature fusion for model construction and training, utilizing the MIMIC dataset. This approach enables us to achieve an accurate estimation of human blood pressure using single PPG as input. Testing the trained model on a dataset that contains more than 60,000 sample segments produced encouraging experimental results. Additionally, through ablation experiments, we substantiated the efficacy of the data generation and feature fusion approaches advanced within this study. Furthermore, using a customized dataset as well as the University of Queensland Vital Signs Dataset, we carried out personalized calibration experiments for each participant. After the calibration procedure, the Mean Absolute Error (MAE) of SBP and DBP showed improvements of 9.60 mmHg and 4.32 mmHg, respectively. The result underscores the practicality of utilizing deep learning for blood pressure estimation while also highlighting the imperative nature of individualized calibration techniques.