What is Rex Cheng's academic and industry background?

Updated

Rex Cheng (Ho Kei Cheng) is a research scientist at Luma AI working on multimodal foundation models. His background combines a PhD at the University of Illinois Urbana-Champaign, earlier study at The Hong Kong University of Science and Technology, and research internships at Adobe Research, Kaiber, Sony AI, and Meta FAIR/MSL.

Current role

  • Research scientist at Luma AI, working on multimodal foundation models.

Education

Stage Institution Advisor(s)
PhD University of Illinois Urbana-Champaign Alexander Schwing
Earlier study The Hong Kong University of Science and Technology Yu-Wing Tai and Chi Keung Tang

Industry research internships

During his PhD, he interned at:

  • Adobe Research
  • Kaiber
  • Sony AI
  • Meta FAIR/MSL

Research output

His site lists selected research with links to project pages, code, and arXiv papers, including:

  • SAM 3: Segment Anything with Concepts (ICLR 2026) — a unified model for detection, segmentation, and tracking of objects in images and video using text, exemplar, and visual prompts.
  • The Curse of Conditions: Analyzing and Improving Optimal Transport for Conditional Flow-Based Generation (ICCV 2025) — condition-aware coupling of prior and data samples for straighter flows.
  • MMAudio: Taming Multimodal Joint Training for High-Quality Video-to-Audio Synthesis (CVPR 2025) — synchronized audio generation from video or text.
  • Putting the Object Back into Video Object Segmentation (CVPR 2024 Highlight) — an object transformer combining pixel-level and object-level features.

The site also links to his GitHub, Google Scholar, and CV for further detail.