- I am a professor at Harbin Institute of Technology (HIT), P. R. China. I had been a distinguished investigator at Xi’an Jiaotong University, P. R. China until Oct. 2021, and a senior researcher/adjunct professor with University of Oulu, Finland until Feb. 2019. I received my Ph.D. degree in computer application and technology from HIT under the supervision of Prof. Wen GAO (FIEEE, FACM), in 2010. I have authored over 50 articles in top-tier publications and conferences such as IEEE T-PAMI, CVPR, ICCV, AAAI, and NeurIPS. I have served as an area chair/SPC for ACM MM 22/21/20, AAAI 22/21/20, and IJCAI22/21, a guest editor for peer-reviewed journals like Patter Recognition Letter and Signal, Image and Video Processing, and a co-organizer for six international workshops in conjunction with IEEE CVPR, ACM MM, IEEE FG. My studies about subtle facial movement analysis have been reported by International media like MIT Technology Review also have been awarded the IEEE Finland Section best student conference paper of 2020. I have been a PI of several projects such as the National Key R&D Program Projects, PRC and Infotech Oulu Postdoctoral funding project, and the project manager of an Academy of Finland ICT 2023 funding project.
- My current research interests include visual surveillance, continual learning, micro-expression analysis, and so forth.
Recent News
- Sept., one paper about domain incremental learning was accepted by NeurIPS 22.
- August, one benchmark paper about deepfake incremental detection was accepted by WACV 23.
- July, one paper about agent based semi-supervised crowd counting was accepted by ACM MM 22.
- April, One joint paper about dual-branch image super-resolution was accepted by IJCAI22.
- April, Github knowledge hub about incremental learning/continual learning was released.
- March, one paper about dynamic local attention for crowd counting got approved by IEEE CVPR 2022. Code
- Feburary, one paper about counting-tracking co-modeling got accepted by IEEE TIP.
- January, one paper about neighborhood maintainance based incremental learning got accepted by IEEE TNNLS.
