Visual perception is an important sensory channel for humans to understand the structure and changes in the external environment. As one of the core areas of artificial intelligence, computer vision aims to utilize principles from cognitive science and computer technology to replicate human visual capabilities. The goal is to enable intelligent perception systems to autonomously understand and interact efficiently with the environment.

The main research topics in computer vision include image recognition and detection, semantic segmentation of images and videos, biometric recognition, 3D scene reconstruction, dynamic and active vision, and human behavior analysis and intention understanding. The primary objective is to equip intelligent perception systems with reliable environmental analysis and understanding capabilities, enabling them to adapt to complex real-world scenarios.

Our research team has undertaken dozens of national, provincial, and industry-funded projects, including the National Natural Science Foundation of China, the 863 Program, the 973 Program, the Ministry of Education's Innovation Team, and the National Key Research and Development Program of China, etc. Our research outcomes have been published in prestigious conferences and journals in the field of computer vision and machine learning, such as TPAMI, IJCV, TVCG, PR, CVPR, ICCV, ECCV, NeurIPS, ICML, AAAI, and ICRA. We have received awards for our achievements, including the National Technology Invention Second Prize, the Ministry of Education Technology Invention First Prize, the Ministry of Public Security Scientific and Technological Progress Second Prize, the Guangzhou Scientific and Technological Progress Third Prize, and the Chinese Academy of Sciences Scientific and Technological Progress Second Prize, and these achievements have been implemented in areas such as criminal investigation, healthcare, and cultural heritage preservation. The Digital Longmen Grottoes project has been featured in mainstream media including CCTV. Our wavelet-based image compression system has been applied in high-speed camera systems in aerospace and digital movie playback systems in cinemas. Our achievements in image understanding, such as multi-view and multi-label learning, have been recognized as key results of the 973 Program. Furthermore, our team have received honors and awards at various academic conferences both domestically and internationally, including the Best Paper/Poster Awards at Euromed 2010, ACPR 2017, CCCV 2017, ICPR 2018, and FG 2020, as well as securing the first place in the V-SLAM at ISMAR 2019.