Personal Information
  • Associate Professor
  • Supervisor of Doctorate Candidates
  • Supervisor of Master's Candidates
  • Name (English):Jifeng DAI
  • E-Mail:
  • School/Department:Dept. EE Tsinghua University
  • Administrative Position:Associate Professor
  • Education Level:With Certificate of Graduation for Doctorate Study
  • Business Address:清华大学罗姆楼8-103
  • Gender:Male
  • Degree:Doctoral Degree in Engineering
  • Status:Employed
  • Alma Mater:清华大学
  • Personal homepage:http://web.ee.tsinghua.edu.cn/daijifeng
Other Contact Information:
  • No content
Profile

Jifeng Dai, Associate Professor, Department of Electronic Engineering, Tsinghua University.

He received his B.S. and Ph.D. degrees from the Department of Automation, Tsinghua University in 2009 and 2014, respectively. His Ph.D. advisor was Professor Jie Zhou. From 2014 to 2019, he worked as a Principal Research Manager in Visual Computing Group at Microsoft Research Asia (MSRA). From 2019 to 2022, he was the Executive Research Director at SenseTime Research. In July 2022, he joined the Department of Electronic Engineering at Tsinghua University as a full-time faculty member. His research focuses on foundation models and core algorithms for visual information understanding. He has achieved internationally leading research results in this field, with over 50,000 citations on Google Scholar.

His major contributions include: (1) Attention mechanisms in visual representations. He developed the Deformable ConvNets series, which have been incorporated into computer vision courses at several top universities and widely adopted in mainstream deep learning frameworks. These models have surpassed those from institutions such as Microsoft, Meta, and Google in performance. (2) Accurate and efficient object detection algorithms. His work on R-FCN has been recognized as one of the “three foundational models for object detection.” His BEVFormer model has attracted significant attention from industry and academic leaders, including NVIDIA CEO Jensen Huang. (3) Design and training of multimodal foundation models. He developed the open-source multimodal foundation model InternVL, which achieves performance comparable to top-tier commercial closed-source models (such as OpenAI's GPT and Google’s Gemini) across more than a dozen key benchmarks, with over 5 million downloads.

He serves as an Associate Editor of TPAMI, an Editorial Board Member of IJCV, and an Area Chair of NeurIPS, ICCV, CVPR, ECCV, and ICLR. He was also the Publicity Chair of ICCV 2019. He leads several high-level national research programs, including the Youth Program of the National Overseas Talent Program, the Young Scientist Project under the Ministry of Science and Technology’s "Next Generation Artificial Intelligence" initiative, a General Program and a Joint Key Program of the National Natural Science Foundation of China (NSFC), and is a core member of an NSFC Innovation Group Project. He is also a recipient of the National Young Talent Program.

Homepage:https://jifengdai.org/

Lab Homepage:https://fundamentalvision.github.io/

Google Scholarhttps://scholar.google.com/citations?user=SH_-B_AAAAAJ&hl=zh-CN

Work Experience
  • Associate Professor , Department of Electronic Engineering , Tsinghua University 2022-7-1 ∼ Now
  • Executive Research Director , SenseTime Research , SenseTime 2019-10-1 ∼ 2022-7-1
  • Principal Research Manager , Visual Computing Group , Microsoft Research Asia 2014-7-1 ∼ 2019-9-1
Educational Experience
  • Tsinghua University,  Department of Automation,  Doctoral degree , With Certificate of Graduation for Doctorate Study 2009-9-1 ∼ 2014-6-1
  • Tsinghua University,  Department of Automation,  Bachelor's Degree , Undergraduate (Bachelor’s degree) 2005-8-1 ∼ 2009-6-1
Social Affiliations
  • Associate Editor of TPAMI
  • Editorial Board Member of IJCV
  • Area Chair of NeurIPS, ICCV, CVPR, ECCV, and ICLR
Research Group
No content
Honored
+

Doctoral Degree in Engineering

Jifeng DAI
MOBILE Version