Hao Zheng
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Hao Zheng has a total of 8 co-authors. Here are the top co-authors by number of shared publications:
- Yihong Chen
Internet of things perception and big data analysis key Laboratory of Nanchong city, China West Normal University, Nanchong, China
2 shared publications - Jiang Zhu
School of Automation and Electronic Information, Xiangtan University, Xiangtan, China
1 shared publication - Meng Liu
China West Normal University
1 shared publication - Ping Huang
Internet of things perception and big data analysis key Laboratory of Nanchong city, China West Normal University, Nanchong, China
1 shared publication - Wanli Ouyang
Shanghai Artificial Intelligence Laboratory
1 shared publication - Yanchun Li
Xiangtan University
1 shared publication
Latest Publications
- Open Access
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 15598-15606
Soundless but Not Lost: Persistent Audio-Spatial Memory for Semantic Audio-Visual Navigation
- 9 November 2026
- Published by Association for Computing Machinery (ACM)
- p. 15598-15606
Open AccessSoundless but Not Lost: Persistent Audio-Spatial Memory for Semantic Audio-Visual Navigation
- 5 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 26 (15209210) , 5769-5781
Online Handwritten Chinese Character Recognition Based on 1-D Convolution and Two-Streams Transformers
As one of the classic problems of pattern recognition, the online Handwritten Chinese Character Recognition (OLHCCR) has attracted the attention of many researchers. Yet, it remains challenging due to complex glyphs, numerous strokes, and huge categories. Existing methods utilize temporal features or spatial features to recognize handwritten characters, which results in recognition errors due to the character with non-standard stroke order. This paper proposes a new OLHCCR model based on 1-D Convolution and Two-Streams Transformers. The model has a 1-D Transformer and a Vision Transformer, and the 1-D Transformer contains a 1-D Convolution layer and Transformers, that is, the model has overall structure of Two-Streams Transformers with 1-D Convolution. So, the model is named as C-TST. It can fuse temporal and spatial features of Chinese character to achieve high recognition accuracy and fast recognition speed. Specifically, each online handwritten Chinese character is represented by a trajectory sequence. The original trajectory sequence is preprocessed to enhance the information density of each trajectory point and features difference among trajectory points. Then, the result after preprocessing is input into the 1-D convolution layer to obtain shallow temporal features, which are used also as the input of the Transformers to capture the temporal features. Simultaneously, character image is generated by processing the original trajectory sequence, and then fed into the Vision Transformer to capture the spatial features. By fusing the captured temporal and spatial features of online handwritten Chinese character, the proposed C-TST achieves a recognition accuracy of 97.90% on ICDAR-2013 and a state-of-the-art recognition accuracy of 97.38% on IAHCC-UCAS2016. The code will be available at https://github.com/cwnuiot/Two-Streams-Transformer
- 5 December 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Multimedia
- Vol. 26 (15209210) , 5769-5781
Online Handwritten Chinese Character Recognition Based on 1-D Convolution and Two-Streams Transformers
As one of the classic problems of pattern recognition, the online Handwritten Chinese Character Recognition (OLHCCR) has attracted the attention of many researchers. Yet, it remains challenging due to complex glyphs, numerous strokes, and huge categories. Existing methods utilize temporal features or spatial features to recognize handwritten characters, which results in recognition errors due to the character with non-standard stroke order. This paper proposes a new OLHCCR model based on 1-D Convolution and Two-Streams Transformers. The model has a 1-D Transformer and a Vision Transformer, and the 1-D Transformer contains a 1-D Convolution layer and Transformers, that is, the model has overall structure of Two-Streams Transformers with 1-D Convolution. So, the model is named as C-TST. It can fuse temporal and spatial features of Chinese character to achieve high recognition accuracy and fast recognition speed. Specifically, each online handwritten Chinese character is represented by a trajectory sequence. The original trajectory sequence is preprocessed to enhance the information density of each trajectory point and features difference among trajectory points. Then, the result after preprocessing is input into the 1-D convolution layer to obtain shallow temporal features, which are used also as the input of the Transformers to capture the temporal features. Simultaneously, character image is generated by processing the original trajectory sequence, and then fed into the Vision Transformer to capture the spatial features. By fusing the captured temporal and spatial features of online handwritten Chinese character, the proposed C-TST achieves a recognition accuracy of 97.90% on ICDAR-2013 and a state-of-the-art recognition accuracy of 97.38% on IAHCC-UCAS2016. The code will be available at https://github.com/cwnuiot/Two-Streams-Transformer
- 15 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 23 (6) , 7301-7317
Body RFID Skeleton-Based Human Activity Recognition Using Graph Convolution Neural Network
Human activity recognition (HAR) has become a hotspot. However, the existing HAR has some shortcomings, such as few recognized human activities, no identification, privacy leakage, and battery maintenance. Therefore, we design body RFID skeleton which fully senses the key features of human activity, and further propose human activity recognitions by taking advantage of RFID (privacy protection, identification, and battery-less maintenance). Firstly, we devise RFID skeleton activity graph as human activity model. Secondly, we propose baseline RFID skeleton activity graph convolution network (FSCN) by using the graph convolution network (GCN), and FSCN classifies the RFID skeleton activity graph for human activity recognition. Thirdly, to solve the tag response signal feature data over-smoothing and the different information aggregation degree in FSCN, we propose improved FSCN with the residual network (R-FSCN). Finally, we design the parallel RFID skeleton activity graph convolution network (PR-FSCN) for optimizing R-FSCN. Massive experiments show that PR-FSCN has comprehensive superiority to the existing HARs. As far as we know, this study is the first work that appropriately designs body RFID skeleton for HAR, and that successfully introduces GCN to investigate the new bound-RFID HAR with high recognition performance. The dataset, source code, and materials are available at http://www.cwnuiot.net/FSCN/ .
- 15 November 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Mobile Computing
- Vol. 23 (6) , 7301-7317
Body RFID Skeleton-Based Human Activity Recognition Using Graph Convolution Neural Network
Human activity recognition (HAR) has become a hotspot. However, the existing HAR has some shortcomings, such as few recognized human activities, no identification, privacy leakage, and battery maintenance. Therefore, we design body RFID skeleton which fully senses the key features of human activity, and further propose human activity recognitions by taking advantage of RFID (privacy protection, identification, and battery-less maintenance). Firstly, we devise RFID skeleton activity graph as human activity model. Secondly, we propose baseline RFID skeleton activity graph convolution network (FSCN) by using the graph convolution network (GCN), and FSCN classifies the RFID skeleton activity graph for human activity recognition. Thirdly, to solve the tag response signal feature data over-smoothing and the different information aggregation degree in FSCN, we propose improved FSCN with the residual network (R-FSCN). Finally, we design the parallel RFID skeleton activity graph convolution network (PR-FSCN) for optimizing R-FSCN. Massive experiments show that PR-FSCN has comprehensive superiority to the existing HARs. As far as we know, this study is the first work that appropriately designs body RFID skeleton for HAR, and that successfully introduces GCN to investigate the new bound-RFID HAR with high recognition performance. The dataset, source code, and materials are available at http://www.cwnuiot.net/FSCN/ .