Hello, I am Jiawei (Phoenix) MA, a final-year CS Ph.D. candidate in DVMM Lab at Columbia University, advised by Prof. Shih-Fu Chang. Before my doctorate program,  I work on computational imaging and am fortunate to be instructed by Prof. Xin Yuan in Westlake University. I can speak English, Chinese and a bit Korean. 

Open Positions (New)

I am going to join the Department of Computer Science at City University of Hong Kong as a Tenure-Track Assistant Professor in 2024 Fall, and building my research team, Statistics & Machine Learning (SMILE) Labortary. For the  academic year of 2024~2025, we have several openings for Ph.D. students and Postdocs, and also welcome various forms of collaboration. Please refer to Labortary and Join Us for detailed information regarding research direction and application instruction.

Research Interest

My research focuses on Machine Learning and Computer vision, with the aim to develop general-purpose intelligent visual system for real-world application. Specifically, my research considers the learning mechanisms for effective and continuous knowledge update with minimal manual curation (e.g., data-centric AI, de-centralized AI), as well as explainable representation that can encode complex raw signals and be robustly applied in diverse tasks & fields (e.g., AI agent, AI for science). 

  • Data-Centric AI: Data mining & Data clustering, Knowledge graph & Sense-making
  • De-Centralized AI: Asynchronous Training, Model Ensemble & Collaboration
  • Representation Learning for Multi-Modal and Dynamic Interaction
  • AI Agent & AI for Science, Data Privacy & Security,


  • [Jun. 2024] Doctoral Consortium CVPR’24 (Incoming), “MoDE: CLIP Data Experts via Clustering”
  • [Mar. 2024] Hong Kong University of Science and Technology (GZ), “Learning with AI Feedback on Data Use”
  • [Feb. 2024] Hong Kong Baptist University, “Learning with AI Feedback on Data Use” [Link]
  • [Feb. 2024] Colorado School of Mines, “Learning with AI Feedback on Data Use”
  • [Jan. 2024] Peking University, “Self-Adaptive Artificial Intelligence via Data-Model Interaction”
  • [Apr. 2023] Virginia Tech, “Efficient Adaptation for Multi-Modal Video Understanding”