Miao Zhang
miaozhng (at) nyu (dot) edu
I am a fourth year Ph.D. candidate at New York University, advised by Prof. Rumi Chunara. I am a member of ChunaraLab and ViDA. Currently, I mainly work on reliable AI: model generalizability and bias mitigation for my Ph.D. degree. I received my Master degree from Stanford University in CCRMA, when I performed research in federated learning for medical image analysis at QIAI advised by Prof. Daniel Rubin. In my undergrad, I worked with Prof. Dong Wang at Tsinghua University on speech processing and recognition.
I have broad interests in:
Responsible AI: Algorithmic fairness, data debiasing, and out-of-domain generalization.
Computer vision and multimodal: Label efficient representation learning.
Health and social science: To adapt pre-trained deep neural networks and LLM for domain specific applications.
news
Sep 18, 2024 | Our paper publishes at Proceedings of the National Academy of Sciences (PNAS). We discuss how big data driven machine learning models can ignore important variables and their effects. We propose a causal framework to improve interpretability and instruct most efficient intervention. |
Aug 13, 2024 | Very excited to present at Cohere for AI about the paper Common-Sense Bias Discovery and Mitigation for Classification Tasks. Check out the online event. |
Jun 1, 2024 | New paper to appear at AAAI/ACM conference on artificial intelligence, ethics, and society 2024. We discuss and promote fairness in self-supervised representation learning for satellite image segmentation between urban and rural areas. |
Jul 1, 2023 | Check out our new project website. In this collaboration with The aga khan university, we collect high resolution satellite imagery to quantify greenspace across city of Karachi, Pakistan. New augmentation strategy integrates domain information and boosts semantic segmentation performance. |
Nov 15, 2022 | I gave a graduate student presentation about my research at Diversity in Research Conference (DIRC 22) |