Junbo Niu

I am a senior student in School of Automation Science, Beihang University. My research focuses on Computer Vision and Multi-Modal Learning including Visual Pretraining, Scene Understanding (Detection and OCR), and Multi-Modal Large Language Models.I am currently involved in an internship at Shanghai AI laboratory, opendatalab, advised by Conghui He . I will pursue a Ph.d's degree in EECS, Peking University starting in Fall 2025, advised by Wentao Zhang and Bin Cui.

Feel free to email me at 21376334@buaa.edu.cn for any form of academic cooperation! More info: 21376334@buaa.edu.cn / Google Scholar / Github / CV

Papers

OVOBench: How Far is Your Video-LLMs from Real-World Online Video Understanding?
Junbo Niu^*, Yifei Li^*, Ziyang Miao, Chunjiang Ge, Jiaqi Wang^† et al.
CVPR 2025
[Paper] [Code]
Supported the development of practical VideoLLMs capable of online processing and response, bridging the gap between model performance and human-level online video understanding in video-based AI.

InternLM-XComposer2.5-OmniLive
Pan Zhang, ...,Junbo Niu, Jiaqi Wang^† et al.
Technical Report
[Paper] [Code]
A Comprehensive Multimodal System for Long-term Streaming Video and Audio Interactions.

Experience

	Shanghai AI laboratory, opendatalab Jul 2024 - Present OCR-Free Visual Understanding && Pretrain Models Research Intern Advisor: Conghui He
	Shanghai AI laboratory, Large Model Center Nov 2023 - May 2024 Video-LLMs && Online Video Understanding Research Intern Advisor: Jiaqi Wang
	Peking University, EECS Starting in Fall 2025 Major in Artificial Intelligence Ph.D. Candidate
	Beihang University, School of Automation Science Sep 2021 - Present GPA 3.85 Rank:1 / 203 B.Eng. Student

Honor

Star of “Yu-Yuan”, Top 2 of 600+ ,The highest honor of the department of Automation.
Undergraduate National Scholarship (rank: 1/156)
Outstanding student of Beihang University(5%)
Merit Student of Beihang University(5%)

Updated on November 12, 2024.