Tianhe Wu

My name is Tianhe Wu, a third-year graduate student at the Intelligent Computing Lab at Tsinghua University, under the supervision of Professor Yujiu Yang. Currently, I am a research assistant at City University of Hong Kong, working with Professor Kede Ma. I am also an intern at the OPPO Research Institute, where I am mentored by Professor Lei Zhang. I am passionate about advancing the frontiers of artificial intelligence.

Email / Google Scholar / Github

Research

My research focuses on Multimodal Large Language Models (MLLMs), Image Quality Assessment (IQA), and Low-level Vision. Currently, most of my work is centered around IQA tasks, with future research directions primarily oriented towards MLLMs.

News

2024-07: Our paper A Comprehensive Study of Multimodal Large Language Models for Image Quality Assessment was accepted at ECCV 2024.
2023-09: Our paper titled Assessor360: Multi-sequence Network for Blind Omnidirectional Image Quality Assessment has been accepted for presentation at NIPS 2023.
2022-04: I published my first conference paper MANIQA: Multi-dimension Attention Network for No-reference Image Quality Assessment.
2022-04: We won first place in the NTIRE2022 Perceptual Image Quality Assessment Challenge, winning both Track 1 (Full-Reference) and Track 2 (No-Reference) competitions.

Publications & Preprints (* means equal contribution)

	A Comprehensive Study of Multimodal Large Language Models for Image Quality Assessment Tianhe Wu, Kede Ma, Jie Liang, Yujiu Yang, Lei Zhang ECCV 2024 [arXiv] [Code] We conducted a systematic evaluation of MLLMs' ability to perceive image quality.
	Assessor360: Multi-sequence Network for Blind Omnidirectional Image Quality Assessment Tianhe Wu, Shuwei Shi, Haoming Cai, Mingdeng Cao, Jing Xiao, Yinqiang Zheng, Yujiu Yang NIPS 2023 [arXiv] [Code] We proposed a new paradram for blind Omnidirectional IQA.
	Attentions Help CNNs See Better: Attention-based Hybrid Image Quality Assessment Network Shanshan Lao, Yuan Gong, Shuwei Shi, Sidi Yang, Tianhe Wu, Jiahao Wang, Weihao Xia, Yujiu Yang CVPRW 2022 Oral [arXiv] [Code] The champion method for NTIRE2022 Full-Reference IQA.
	MANIQA: Multi-dimension Attention Network for No-Reference Image Quality Assessment Sidi Yang, Tianhe Wu, Shuwei Shi, Shanshan Lao, Yuan Gong, Mingdeng Cao, Jiahao Wang, Yujiu Yang CVPRW 2022 Oral [arXiv] [Code] The champion method for NTIRE2022 No-Reference IQA.