Home

Greetings! I am a Ph.D. student at Language Technologies Institute, School of Computer Science, Carnegie Mellon University, working with Prof. Alexander G. Hauptmann. I am also a Student Researcher at Google. I graduated summa cum laude from Peking University, China, with double bachelor’s degrees in Computer Science and Economics. Here is my Curriculum Vitae.

My research interests include content creation, multimedia, and video understanding. I am currently working on video generation, see the latest release: MAGVIT.

News

• [02/2023] One paper for transformer-based video generation accepted at CVPR 2023 as a highlight (top 2.5% among 9.2k submissions).
• [01/2023] One paper for continuous-time discrete diffusion accepted at ICLR 2023.
• [11/2022] We introduce the multi-task masked generative video transformer, MAGVIT.
• [07/2022] One paper for zero-shot action recognition accepted at ECCV 2022.
• [11/2021] We helped the Washington Post in analyzing the crowd density at the Astroworld Festival, watch.
• [11/2021] We won the 1st place at MediaEval 2021 Sports Video Task: Stroke Classification.
• [10/2021] We won the 1st place at ICCV 2021 ROAD Challenge: Action Detection Task.
• [06/2021] We won the 1st place at CVPR 2021 ActivityNet Challenge: ActEV SDL and Kinetics-700 tasks.
• [01/2021] We introduce the Argus Activity Detection System, watch:
• [01/2021] We won the 1st place at WACV 2021 ActEV SDL Unknown Facilities Challenge.
• [11/2020] We won the 1st place at NIST TRECVID 2020 ActEV Evaluation.
• [09/2020] I was named 2021 Siebel Scholars with a $35k fellowship, top 5 and the only Master’s student at CMU SCS. Press: BusinessWire, Bloomberg, Yahoo, CMU.
• [04/2020] We won the 1st place at CVPR 2020 AI City Challenge: City-Scale Multi-Camera Vehicle Tracking with a prize of NVIDIA Quadro GV100.

Selected Publications (Full List)

MAGVIT: Masked Generative Video Transformer

Lijun Yu, Yong Cheng, Kihyuk Sohn, José Lezama, Han Zhang, Huiwen Chang, Alexander G Hauptmann, Ming-Hsuan Yang, Yuan Hao, Irfan Essa, Lu Jiang. In CVPR Highlight, 2023

Score-based Continuous-time Discrete Diffusion Models

Haoran Sun, Lijun Yu, Bo Dai, Dale Schuurmans, Hanjun Dai. In ICLR, 2023

Rethinking Zero-shot Action Recognition: Learning from Latent Atomic Actions

Yijun Qian, Lijun Yu, Wenhe Liu, Alexander G. Hauptmann. In ECCV, 2022

Argus++: Robust real-time activity detection for unconstrained video streams with overlapping cube proposals

Lijun Yu, Yijun Qian, Wenhe Liu, Alexander G. Hauptmann. In WACVW, 2022

Zero-VIRUS: Zero-shot VehIcle Route Understanding System for Intelligent Transportation

Lijun Yu, Qianyu Feng, Yijun Qian, Wenhe Liu, Alexander G. Hauptmann. In CVPRW, 2020

Selected Talks (Full List, Talk Map)

Argus++: Real-time Activity Detection in Unknown Facilities with Dense Spatio-temporal Proposals

June 18, 2021
Invited Talk at ActivityNet Challenge, IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Virtual online

Argus: Efficient Activity Detection System for Extended Video Analysis

March 05, 2020
Invited talk at Workshop on Human Activity Detection in multi-camera, Continuous, long-duration Video (HADCV 20), 2020 IEEE Winter Conference on Applications of Computer Vision (WACV), Aspen, Colorado

Traffic Danger Recognition With Surveillance Cameras Without Training Data

November 13, 2019
Invited Talk at NIST TREC Video Retrieval Evaluation (TRECVID) 2019, Gaithersburg, Maryland

Selected Projects (Full List)

Rethinking Zero-shot Action Recognition: Learning from Latent Atomic Actions

Pyturbo Package: A Pipeline System for Efficient Concurrent Execution

Zero-VIRUS: Zero-shot VehIcle Route Understanding System for Intelligent Transportation (CVPR 2020 AI City Challenge Track 1)