Scholar
Haodong Duan
Google Scholar ID: vi3W-m8AAAAJ
Shanghai AI Lab | CUHK | PKU
Citations & Impact
Citations
7,554
H-index
31
i10-index
47
Publications
20
Co-authors
39
list available
Publications
Resume (English only)
Academic Achievements
- Three papers accepted by NeurIPS 2024 main conference: InternLM-XComposer2-4KHD, MMStar, Prism (September 2024).
- Three papers accepted by NeurIPS 2024 Dataset & Benchmark Track: ShareGPT4Video, GMAI-MMBench, MMBench-Video (September 2024).
- MMBench accepted by ECCV 2024 as an oral presentation (August 2024).
- MathBench accepted by ACL 2024 (May 2024).
- Two papers (BotChat, Ada-LEval) accepted by NAACL 2024 (March 2024).
- Released VLMEvalKit, an all-in-one toolkit for evaluating LVLMs, and it was accepted by MM 2024 (December 2023).
- SkeleTR accepted by ICCV 2023 (October 2023).
- Released PYSKL, a codebase for skeleton action recognition, and it was accepted by MM 2022 (May 2022).
- Three papers accepted by CVPR 2022, with PoseC3D and TransRank as oral presentations and OCSampler as a poster (March 2022).
- OmniSource accepted by ECCV 2020 (July 2020).
- TRB accepted by ICCV 2019 as an oral presentation (July 2019).
Research Experience
- Joined Shanghai AI Lab as a postdoctoral researcher in October 2023.
- Interned at AWS AI from July 2022, advised by Dr. Mingze Xu.
- Joined OpenMMLab in August 2020 and served as a maintainer of MMAction2.
- Served as a reviewer for multiple international conferences such as ICCV, AAAI, CVPR, ECCV, NeurIPS, etc.
- Acted as a reviewer for several journals including TPAMI, IJCV, TIP, etc.
Background
- His research interests include video recognition, human-centric action understanding, and multi-modality learning. He is currently a postdoctoral researcher at Shanghai AI Lab, focusing on the evaluation of large language models and multi-modality models.
Miscellany
- His homepage provides a link to download his CV, and his team is hiring full-time researchers/engineers and interns.
