The Vision & Security Laboratory (VSLab) is affiliated to the National Laboratory of Pattern Recognition( NLPR), Institute of Automation, Chinese Academy of Sciences( CASIA). VSLab is led by Prof. Weiming Hu, with an energetic team consisting of 15 full-time researchers and nearly 30 MSc/PhD candidates. The research interests of VSLab mainly cover computer vision, pattern recognition, and machine learning, with a focus on the following topics: object tracking, object detection & segmentation, action recognition, face forgery detection, knowledge distillation, multi-label recognition, multiple instance learning, weakly-supervised learning, etc. Until now, VSLab has published over 200 papers, in which more than 100 papers are on top journals or conferences (TPAMI / IJCV / CVPR / ICCV / ECCV, etc). Besides, VSLab has won multiple top academic awards, e.g., the Visual-Object-Tracking 2018 (VOT 2018) real-time track and ICCV 2019 - English & Chinese double tracks of VATEX Captioning Challenge. In short, VSLab is devoted to the research of fundamental theories and innovative methods around “AI + Security”.


  • 10/2020: 3rd Chinese Conference on Pattern Recognition and Computer Vision (PRCV 2020)  best paper award.
  • 08/2020: Two papers on visual tracking and saliency prediction are accepted by ECCV 2020.
  • 06/2020: Two papers on visual tracking and video captioning are accepted by CVPR 2020.
  • 12/2019: This year we have one paper accepted by IJCV, one paper accepted by TIP and one paper accepted by PR.
  • 11/2019: Our team wins VATEX Captioning Challenge 2019.
  • 06/2019: Three papers on visual tracking and knowledge distillation are accepted by CVPR 2019.
  • 12/2018: This year we have two papers accepted by TPAMI and two papers accepted by TIP.
  • 09/2018: Our team wins VOT2018 real-time challenge.
  • 09/2018: Three papers on visual tracking and action recognition are accepted by ECCV 2018.
  • 07/2018: One paper on visual tracking is accepted by IJCAI 2018.
  • 06/2018: Two papers on visual tracking and facial age estimation are accepted by CVPR 2018.
  • 02/2018: One paper on action recognition is accepted by AAAI 2018.
  • 12/2017: This year we have two papers accepted by TPAMI and two papers accepted by PR.
  • 10/2017: One paper on visual object tracking is accepted by ACM MM 2017.
  • 07/2017: One paper on dynamic object detection is accepted by CVPR 2017.