Gongye Liu
Scholar

Gongye Liu

Google Scholar ID: AWngqwgAAAAJ
The Hong Kong University of Science and Technology
Image/Video Generation
Citations & Impact
All-time
Citations
644
 
H-index
7
 
i10-index
7
 
Publications
9
 
Co-authors
9
list available
Resume (English only)
Academic Achievements
  • - Publications:
  • * Flow-GRPO: Training Flow Matching Models via Online RL, NeurIPS 2025
  • * Improving Video Generation with Human Feedback, NeurIPS 2025
  • * Autoregressive Models in Vision: A Survey, TMLR 2025
  • * StyleCrafter: Taming Stylized Video Diffusion with Reference-Augmented Adapter Learning, TOG 2024
  • * DynamiCrafter: Animating Open-domain Images with Video Diffusion Priors, ECCV 2024
  • * Accelerating Diffusion Models for Inverse Problems through Shortcut Sampling, IJCAI 2024
Research Experience
  • - KwaiVGI, Kuaishou, Research Intern, focus on RLHF in Vision Generation, Mentor: Dr. Menghan Xia, Jiajun Liang, and Dr. Xintao Wang, Sep. 2024 - Present
  • - Hunyuan Video Generation Center, Tencent, Engineering Intern, focus on Human-Centered AIGC, Mentor: Hongfa Wang, Apr. 2024 - Aug. 2024
  • - AILab CVC, Tencent, Research Intern, focus on Controllable Video Generation, Mentor: Dr. Menghan Xia, Jul. 2023 - Mar. 2024
Education
  • - The Hong Kong University of Science and Technology, Ph.D. Student in Individualized Interdisciplinary Program (IIP), Supervisor: Prof. Wenhan Luo, Sep. 2025 - Present
  • - Tsinghua University, Master Degree in Artificial Intelligence, Supervisor: Prof. Yujiu Yang, Sep. 2022 - Jun. 2025
  • - Beihang University, Bachelor Degree in Aircraft Control and Information Engineering, Sep. 2018 - Jun. 2022
Background
  • - Research Interests: Generative AI, with a focus on diffusion/flow-matching models for images and videos
  • - Professional Field: Foundation models, controllable generation, and post-training techniques like RL
Miscellany
  • - Personal Interests: Not provided