alt text  

Huaibo Huang (黄怀波)  

Associate Professor, MAIS & NLPR, CASIA
University of Chinese Academy of Sciences

Google Scholar  |  GitHub
Email: huaibo.huang@cripac.ia.ac.cn


Biography

I am an associate professor at Institute of Automation, Chinese Academy of Sciences (CASIA). I obtained my Ph.D. degree from CRIPAC-NLPR-CASIA in 2019, under the supervision of Prof. Tieniu Tan. Prior to that, I received my M.E. from Beihang University in 2016, supervised by Prof. Di Feng, and my B.E. from Xi’an Jiaotong University in 2012.


My current research interests mainly focus on multimodality, generative learning and low-level vision.


I am open to any discussion or collaboration. If you are interested, please feel free to contact me.

Recent Publications

alt text  

Rectifying Magnitude Neglect in Linear Attention
Qihang Fan, Huaibo Huang, Yuang Ai, Ran He
ICCV 2025 (Highlight) Code

alt text  

Semantic Equitable Clustering: A Simple and Effective Strategy for Clustering Vision Tokens
Qihang Fan, Huaibo Huang, Mingrui Chen, Ran He
ICCV 2025

alt text  

Breaking the Low-Rank Dilemma of Linear Attention
Qihang Fan, Huaibo Huang, Ran He
CVPR 2025 Code

alt text  

MMFakeBench: A Mixed-Source Multimodal Misinformation Detection Benchmark for LVLMs
Xuannan Liu, Zekun Li, Pei Pei Li, Huaibo Huang, Shuhan Xia, Xing Cui, Linzhi Huang, Weihong Deng, Zhaofeng He
ICLR 2024 Project Code

alt text  

DreamClear: High-Capacity Real-World Image Restoration with Privacy-Safe Dataset Curation
Yuang Ai, Xiaoqiang Zhou, Huaibo Huang, Xiaotian Han, Zhengyu Chen, Quanzeng You, Hongxia Yang
NeurIPS 2024 Code

alt text  

Visual Anchors Are Strong Information Aggregators For Multimodal Large Language Model
Haogeng Liu, Quanzeng You, Xiaotian Han, Yongfei Liu, Huaibo Huang, Ran He, Hongxia Yang
NeurIPS 2024 Code

alt text  

Hallo3D: Multi-Modal Hallucination Detection and Mitigation for Consistent 3D Content Generation
Hongbo Wang, Jie Cao, Jin Liu, Xiaoqiang Zhou, Huaibo Huang, Ran He
NeurIPS 2024

Academic Services

Honors and Awards