Guangxi Key Laboratory of Manufacturing System and Advanced Manufacturing Technology, School of Mechanical Engineering, Guangxi University, Nanning, China
Auto-generated profile, based on publicly available information
Yingzhe Tang
Guangxi Key Laboratory of Manufacturing System and Advanced Manufacturing Technology, School of Mechanical Engineering, Guangxi University, Nanning, China
Data glove is one of the devices that can express human intention via detecting hand motion with high-precision. In recent years, flourished applications have been driven by data gloves from human-machine interaction (HMI) to biomedical sciences. To understand the latest research status and grasp the development direction, this study provides a systematic overview of the state-of-the-art for data gloves. Specifically, we take the research topics of hardware design, algorithms for gesture recognition, and the application areas as the entry point. Detailed elaboration of key algorithms techniques, from data processing to classification, for hand gesture recognition is made and quantitative analysis of the recent research achievements is summarized. We found that hand motion detection with data gloves is mostly employed for gesture recognition, and the majority of studies are focused on the performance validation of the gesture recognition, in particular for static gestures. A systematic illustration of the current research challenges as well as the future directions of related data gloves are finally discussed.
13 May 2024 in IEEE Transactions on Instrumentation and Measurement
Data glove-based dynamic gestures contain rich human motion intentions, which are reliant on the hand–body information that comes from multiindividual sensors attached. However, present gesture recognition with such wearable sensor devices tends to depend heavily on handcrafted features and ignore the critical channel and interfeature information. To address this problem, a novel convolutional-transformer-based recognition architecture termed the spatial–temporal feature-attention transformer network (STFTnet) is proposed in this study. Specifically, the acquired data from multiple sensors of the data glove are sequentially processed with a spatial–temporal sensor features embedding branch, a transformer encoder block, and the final gesture classifier. A multisensor feature attention (MFA) block and an improved depth-separable convolution block of the first branch are developed to effectively extract low-level spatial and local temporal features, while the multihead self-attention-based transformer block further concentrates on capturing the global context information. The gesture classifier is used to achieve the final classification successfully. To evaluate the efficacy of the proposed approach, extensive experiments are conducted on two publicly available datasets of pelvic closed reduction action dataset and the UC2017 Hand Gesture (HG) Dataset, and one self-built gesture control command dataset. Compared to the other state-of-the-art deep learning (DL)-based algorithms, an average accuracy of 95.75%, 100%, 99.72% and recognition time of 10.71, 11.92, and 11.24 ms has been achieved. These results indicate that the proposed network effectively enhances the recognition performance of the dynamic gesture of data gloves, while fulfilling requirements of the further real-time application.