About Me

I’m a Senior Applied Scientist at Amazon AGI, where I work on building Amazon’s foundation models. I’m a key contributor to the Amazon Nova family of models, including leading the vision modality for Nova Multimodal Embeddings and the video pretraining data for Nova 2.0. Prior to joining AGI, I worked at Amazon Prime Video on machine learning models for video understanding.

I received my Ph.D. in Computer Science (2022) from Arizona State University, supervised by Prof. Fengbo Ren, and my B.S. in Statistics (2016) from the University of Science and Technology of China.

My research spans multimodal learning, vision-language models, video understanding, and contrastive learning. I have published at top venues including NeurIPS, ECCV, WACV, ICASSP, and IROS, and hold six granted patents. I serve as a peer reviewer for CVPR, ICCV, ECCV, NeurIPS, AAAI, WACV, and ACL, and was recognized as an Outstanding Reviewer (top 5%) at CVPR 2025.