Bowen Cheng 程博文

Bowen is a researcher at TBD Lab, Meta.

Bowen was previously a researcher at OpenAI, working on multimodal post-training (perception and interaction). Before that, Bowen was a researcher at Tesla. Bowen received his Ph.D. at the University of Illinois Urbana-Champaign (UIUC). His Ph.D. advisors are Prof. Alexander Schwing and Prof. Thomas Huang (2017-2020).

Research Interests

I believe multimodal is the core interaction layer between human and AI. (1) perception is the foundation to grounds everything that follows (2) interaction is the key to makes human-AI communication natural and intuitive (3) generation is the cherry on top that makes information easier to understand and consume

Selected projects

• Thinking with Images - initiate research and foundational contributor
• OpenAI o3 and o4-mini - core contributor
• GPT-4.1 - core contributor
• OpenAI Audio API - research
• GPT-4o - core contributor (perception and advanced voice mode)
• Tesla FSDv12 - core contributor
• Mask2Former
• MaskFormer
• Panoptic-DeepLab

Please refer to my Google scholar for a full list of my publications.