I am a 4th-year Ph.D candidate at University of Science and Technology of China (USTC), under the supervision of Dong Liu.

I’m generally interested in video self-supervised learning. In particular, on topics:

video correspondence learning with downstream tasks including video object segmentation, point tracking, etc.
learning generalizable representations with large-scale unlabled videos
large-scale generative models for representation and synthesis, e.g., diffusion models, Masked Image Modeling.

🔥 News

📝 Publications

ICCV 2023

Learning Fine-Grained Features for Pixel-wise Video Correspondences [pdf][code]

Rui Li, Shenglong Zhou, and Dong Liu

IEEE/CVF International Conference on Computer Vision (ICCV) 2023

Learn the fine-grained features for pixel-level video correspondences by training on a mixture of labeled synthetic data and unlabeled real data

CVPR 2023

Spatial-then-Temporal Self-Supervised Learning for Video Correspondence [pdf][code]

Rui Li, Dong Liu

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR) 2023

Propose a spatial-thentemporal self-supervised learning method to find video correspondence

ICCV 2021

Motion-Focused Contrastive Learning of Video Representations [pdf][code]

Rui Li, Yiheng Zhang, Zhaofan Qiu, Ting Yao, Dong Liu, and Tao Mei

IEEE/CVF International Conference on Computer Vision (ICCV) 2021

Leverage motion information to boost self-supervised video representation learning
Selected as Oral Presentation (top 3%) on ICCV 2021

ACM MM 2019

BERT4SessRec: Content-based video relevance prediction with bidirectional encoder representations from transformer [pdf] [code]

Xusong Chen, Dong Liu, Chenyi Lei, Rui Li, Zheng-Jun Zha, and Zhiwei Xiong

ACM Multimedia (MM) 2019

Reviewer of CVPR 2023, CVPR 2022, and TMM.