Research Paper by ISD Undergraduate Accepted to Top-Tier Conference in Computer Vision
Mr. XU Yinzhe (Shaun) our undergraduate student, has made significant research findings in the field of computer vision. Under the supervision of Prof. Sai-Kit Yeung, Shaun and his team developed a new benchmark dataset and framework for omnidirectional visual object tracking, which has been accepted by the top-tier conference International Conference on Computer Vision (ICCV) 2023.
Visual object tracking is an important problem in computer vision with applications in video analysis, human-machine interaction, and intelligent robots. Shaun and his team found that there has been little attention paid to omnidirectional visual object tracking, although its advantages, such as stable long-term tracking and perception, are crucial in many applications. To address this gap, they proposed a new large-scale omnidirectional tracking benchmark dataset, 360VOT, with a new general tracking framework dedicated to this subfield to facilitate future research. To enable accurate performance evaluation in the subfield, they also provided four types of unbiased ground truth, including (rotated) bounding boxes and (rotated) bounding field-of-views, as well as new metrics tailored for omnidirectional images.
Shaun will continue his studies at HKUST as a new ISD PhD student in Prof. Sai-Kit Yeung's team this year. For more information about their research, please visit the website for 360VOT: https://360vot.hkustvgd.com.