YUAN Li

YUAN Li is an Assistant Professor in the School of Electronic and Computer Engineering at Peking University. His research interests include computer vision, deep neural network architecture design, and machine learning. His work has been published in top computer conferences and journals such as CVPR, ICCV, ICML, NeurIPS, AAAI, and IEEE TPAMI. He is an Associate Editor of IEEE IROS 2022 and a member/reviewer of the (senior) Program Committee for several conferences and journals. Dr. Yuan's academic work includes deep neural network frameworks such as VOLO, T2T-VIT, and knowledge distillation. He was a visiting researcher at Harvard University during his PhD.

Li Yuan is a researcher specializing in multimodal machine learning and computer vision. Their work focuses on advancing vision transformers, visual representation learning, and large vision-language models. Key contributions include developing methods for training vision transformers from scratch, improving video understanding through alignment techniques, and enhancing image and video retrieval with efficient quantization strategies. Li Yuan's research also explores self-supervised learning for point clouds, knowledge distillation for object detectors, and the integration of spiking neural networks with transformer architectures. Their publications highlight efforts to unify visual and language understanding through multimodal pretraining and to improve the reasoning capabilities of vision-language models.