About Me

I am a Research Scientist at NVIDIA Cosmos Lab, where I work on world foundation models and physical AI. I received my Ph.D. in Computer Science from Brown University, advised by Prof. Chen Sun, and my bachelor's degree in software engineering from Tsinghua University.

My research interests involve building physically grounded, reasoning-capable vision-language and world models, and exploring how they can be integrated into the physical world. Previously, I worked as a research intern at Salesforce AI Research, Meta GenAI, Google DeepMind, and Google Research. Feel free to contact me for collaborations and casual chats.

Selected Publications

• Active Video Perception: Iterative Evidence Seeking for Agentic Long Video Understanding [paper] [website]
  Ziyang Wang, Honglu Zhou, Shijie Wang, Junnan Li, Caiming Xiong, Silvio Savarese, Mohit Bansal, Michael S. Ryoo, Juan Carlos Niebles
  CVPR Findings 2026
• MotiF: Making Text Count in Image Animation with Motion Focal Loss [paper] [website] [benchmark]
  Shijie Wang, Samaneh Azadi, Rohit Girdhar, Sai Saketh Rambhatla, Chen Sun, and Xi Yin
  CVPR 2025
• How Can Objects Help Video-Language Understanding? [paper]
  Zitian Tang, Shijie Wang, Junho Cho, Jaewook Yoo, and Chen Sun
  ICCV 2025
• Learning Visual Grounding from Generative Vision and Language Model [paper]
  Shijie Wang, Dahun Kim, Ali Taalimi, Chen Sun, and Weicheng Kuo
  WACV 2025
• Vamos: Versatile Action Models for Video Understanding [paper] [website] [code]
  Shijie Wang, Qi Zhao, Minh Quan Do, Nakul Agarwal, Kwonjoon Lee, and Chen Sun
  ECCV 2024
• AntGPT: Can Large Language Models Help Long-term Action Anticipation from Videos? [paper] [website] [code]
  Qi Zhao*, Shijie Wang*, Ce Zhang, Changcheng Fu, Minh Quan Do, Nakul Agarwal, Kwonjoon Lee, and Chen Sun
  ICLR 2024

Experience

Education

Service

Reviewer: TPAMI, IJCV, ICLR, ICML, NeurIPS, CVPR, ICCV, ECCV, AAAI, and WACV.