Ming-Chang Chiu

About

I am a researcher working at the intersection of trustworthy AI, vision-language models, and agentic systems. My work focuses on building robust and aligned multimodal systems that can reason, verify, and collaborate. I am recently facinated by the design of mechanism and infrastructure of Agent-Agent interactions.

I am a PhD Candidate advised by Prof. Xuezhe Ma at University of Southern California (USC) and I work very closely with Dr. Pin-Yu Chen at IBM Research. During my PhD, I have been fortunate to work at NVIDIA, Google Research (now DeepMind), Lawrence Livermore National Lab, Lilt and Illumina.

Selected Publications

Google Scholar

AIDE: Agentically Improve Visual Language Model with Domain Experts ICLR 2025 Self-Improving Foundation Model
MC Chiu*, F Liu, K Sapra, A Tao, Y Jacoob, X Ma, Z Yu, G Liu [PDF]

ColorSense: A Comparative Study on Color Vision of Machine Visual Recognition Secure and Trustworthy ML 2025
MC Chiu*, Y Wang, DEG Kim, PY Chen, X Ma [PDF] [Slides] [Poster]

VideoPoet: A Large Language Model for Zero-Shot Video Generation ICML 2024 (Best Paper)
Kondratyuk et al. [PDF] [Demo]

MegaCoin: Medium Granularity Color Annotated Images Improve Instruction Tuning for Vision-Language Models ArXiv
MC Chiu*, S Wen, PY Chen, X Ma [PDF]

Behavioral Bias of Vision-Language Models: A Behavioral Finance View ICML 2024 LLMs and Cognition
Yuhang Xiao, Yudi Lin, Ming-Chang Chiu* [PDF] [Poster]

DDI-CoCo: Understanding the Effect of Color Contrast in Machine-Assisted Skin Disease Detection ICASSP 2024
MC Chiu*, Y Wang, YJ Kuo, PY Chen [PDF] [Poster]

Better may not be fairer: A study on subgroup discrepancy in image classification ICCV 2023
MC Chiu*, PY Chen, X Ma [PDF] [Poster]

Services

Reviewer for: ICML, NeurIPS, ACL, ICASSP, IJCNN

Teaching: DSCI-553 Foundations and Applications of Data Mining, DSCI-552 Machine Learning for Data Science, CSCI-544 Applied Natural Language Processing

About

CV

Selected Publications

Services

MISC

Contact