Scholar
Pengfei Wan
Google Scholar ID: P6MraaYAAAAJ
Head of Kling Video Generation Models, Kuaishou Technology
Citations & Impact
Citations
3,548
H-index
27
i10-index
45
Publications
20
Co-authors
41
list available
Publications
Resume (English only)
Academic Achievements
- Published multiple papers in top international conferences and journals such as NeurIPS, SIGGRAPH Asia, ICCV, CVPR, ICML Spotlight, ICLR, TPAMI, ECCV, including Flow-GRPO, OmniSync, Context as Memory, FullDiT, ReCamMaster, Towards Precise Scaling Laws for Video Diffusion Transformers, MODA, 3DTrajMaster, DVIS++, Agent Attention, VideoTetris, I2V-Adapter, Augmentation-Aware Self-Supervision for Data-Efficient GAN Training, Towards Practical Capture of High-fidelity Relightable Avatars, FEditNet, etc.
Research Experience
- Served as the head of the Visual Generation and Interaction Center at Kuaishou Technology and the director of AI department at Meitu, Inc.
Education
- Ph.D. from the ECE Department of HKUST; B.E. from the EEIS Department of USTC.
Background
- Head of the Visual Generation and Interaction Center (aka Kling Team) at Kuaishou Technology. Former director of AI department (MT Lab) at Meitu, Inc. Passionate about building intelligent multimodal content creation algorithms and systems, both for people (to express and live better) and for machines (to simulate and evolve better). Recent focus on video generation, multimodal interaction, and related topics.
Miscellany
- Interested in computer vision, graphics, generative AI, and multimodal machine learning.
