Wei-Jyun Chen

Wei-Jyun Chen

Auto-generated profile, based on publicly available information

Wei-Jyun Chen

Department of Electrical Engineering, Institute of Communications Engineering, National Tsing Hua University, Hsinchu, Taiwan

Research Focus & Metrics

Academic Career Length

16Years (2011-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
22 July 2022 in IEEE Transactions on Circuits and Systems I: Regular Papers
Verified
Journal Article
Closed access
CCChing-Te Chiu
Ching-Te Chiu
YDYu-Chun Ding
Yu-Chun Ding
WLWei-Chen Lin
Wei-Chen Lin
WCWei-Jyun Chen
Wei-Jyun Chen
SWShu-Yun Wu
Shu-Yun Wu
YHYuan-Hao Huang
Yuan-Hao Huang
Face classification is important in many applications such as surveillance, border control, and security systems. However, wide variations in environments such as insufficient light, large distances or pose angles make the task challenging. Depth sensors are added with RGB cameras for improving classification accuracy but commercial RGB-D sensors are most targeted for indoors applications. In this paper, we present and design a Chaos LiDAR depth sesnor that provides high-precision depth images through intelligent correlation processing for both indoors and outdoors applications. Our Chaos LiDAR depth sensor detects range from 2 to 40 meters with precision around 8mm at 20-meter. With the Chaos LiDAR depth as input, we design a RGB-D based face classification embedded CNN (eCNN) model for wide range applications such as dim illumination, various distances and large poses. Our Chaos LiDAR increases around 14.27% classification accuracy compared to RealSense D435i for distance from 3 to 5 meter. The eCNN face classification subsystem is implemented in Xilinx ZCU 102 and achieves 11.11 ms inference time. The eCNN engine achieves a peak throughput at 614.4 GOPS. The overall system including Chaos LiDAR, correlation and eCNN FPGA achieves face classification inference rate of 10fps.

1 November 2011
Verified
Conference Paper
Closed access
KSKai-Tai Song
Kai-Tai Song
WCWei-Jyun Chen
Wei-Jyun Chen
This paper presents a vision-based human activity recognition system using a mobile camera. This system aims to enhance human-robot interaction in a home setting for applications such as health care and companion. In the first place, the camera needs to find a human in image frames. The body pose is classified for the detected human. Then the human activity is recognized by combining information of human pose, human location and elapsed time. In order to determine the situated place of the person in a home setting, a novel space-boundary detection method is proposed in this paper. This method uses features in the environment to automatically set space boundary in the image such that human location in the environment can be obtained. In the integrated experiments, human pose recognition rate of five poses(standing, walking, sitting, squatting, lying) is 94.8%. Experiments of human activity recognition in a home setting have been conducted to verify the performance of the proposed method by using a mobile camera from different view angles and positions in a home setting. The experimental results reveal that the space boundaries are detected as expected and satisfactory results are obtained.