Jae Sung (James) Park
Scholar

Jae Sung (James) Park

Google Scholar ID: hD2WqqcAAAAJ
University of Washington
Computer VisionNatural Language ProcessingMachine Learning
Citations & Impact
All-time
Citations
1,680
 
H-index
12
 
i10-index
13
 
Publications
18
 
Co-authors
0
 
Publications
18 items
Browse publications on Google Scholar (top-right) ↗
Resume (English only)
Academic Achievements
  • Published two papers, 'Molmo' and 'Synthetic Visual Genome', at CVPR 2025, one of which is a Best Paper Award Candidate; Published one paper at Neurips and one at Neurips D&B in 2024; Released Molmo, an open state-of-the-art multimodal AI model; Other notable works include BLIP3-KALE, Certainly Uncertain benchmark, etc.
Research Experience
  • Internship at Microsoft Research Deep Learning Team (Spring 2022 - Winter 2024); Co-organized ECCV2024 Workshop on Multimodal Agents and CVPR 2024 Tutorial on Generalist Agent AI.
Education
  • PhD in Computer Science and Engineering from University of Washington, advised by Yejin Choi, Ali Farhadi, and Ranjay Krishna; B.S. in EECS from University of California, Berkeley, worked closely with Anna Rohrbach and Trevor Darrell.
Background
  • Research Interests: visual perception and language understanding, specifically how machines can reason about the visual world as humans do. Research projects focused on: Empowering Visual Commonsense Reasoning of AI models, Grounding Objects, Concepts, Actions to Images and Videos, Evaluation of Multimodal Language Models.
Miscellany
  • Contact information includes Email, Google Scholar, and GitHub.
Co-authors
0 total
Co-authors: 0 (list not available)