Qinghua Hu
Also known as 胡清华, Hu Qing-Hua, Hu Qing-hua, Hu QingHua, Hu Qinghua, Hu qinghua
Qinghua Hu is a machine-learning researcher at Tianjin University whose work spans multi-view clustering and deep-learning attention. He earned three degrees at Harbin Institute of Technology through 2008, held a postdoctoral fellowship at Hong Kong Polytechnic University, and became a full professor at Tianjin University in 2012. At CVPR, he published a 2017 paper on latent multi-view subspace clustering and coauthored ECA-Net, a 2020 channel-attention module for convolutional neural networks. He received the Wu Wenjun Artificial Intelligence Science and Technology Progress First Prize in 2024. Since May 2025, he is Associate Dean of Tianjin University's Graduate School and Dean of its College of Artificial Intelligence, and directs the Tianjin Key Laboratory of Machine Learning.
- Now at
- Tianjin University
- Works
- 684
- Citations
- 40,746
- h-index
- 87
Publication numbers from OpenAlex, as of 22 October 2026.
Timeline
-
1995–2008
Degrees at Harbin Institute of Technology
He earns a bachelor's degree in the School of Energy Science and Engineering at Harbin Institute of Technology in 1999 and a master's degree there in 2002. He completes a PhD through its School of Aerospace in 2008.
-
2009–2011
Postdoctoral work in Hong Kong
He becomes a postdoctoral fellow in the Department of Computing at Hong Kong Polytechnic University in October 2009 and serves through October 2011.
-
2012
Professor at Tianjin University
In April 2012, he becomes a full professor in Tianjin University's School of Computer Science and Technology.
-
2016
Heilongjiang Natural Science First Prize
He receives the Heilongjiang Provincial Natural Science First Prize in 2016 for work on robust, data-driven modeling of power-system fault diagnosis.
-
2017
Latent multi-view clustering
In 2017, he coauthors Latent Multi-View Subspace Clustering at CVPR. The method clusters points through a latent representation and uses complementary information across views to make the representation more comprehensive, accurate and robust.
-
2020
ECA-Net at CVPR
With Qilong Wang and colleagues, he coauthors ECA-Net at CVPR 2020, an efficient channel-attention module for convolutional neural networks. It uses one-dimensional convolution for local cross-channel interaction without dimensionality reduction.
-
2024
Wu Wenjun AI award
He receives the Wu Wenjun Artificial Intelligence Science and Technology Progress First Prize in 2024 for key technologies and applications in low-altitude intelligent sensing.
-
Since May 2025
Dean of Tianjin's AI College
In May 2025, he becomes Associate Dean of Tianjin University's Graduate School and Dean of its College of Artificial Intelligence. In 2026, he directs the Tianjin Key Laboratory of Machine Learning.
Career
- Since 2012Tianjin UniversityFull Professor, Associate Dean
- 2025–2026Soochow University
- 2023–2025Qufu Normal University
- 2023–2026State Council of the People's Republic of China
- 2019–2023Chinese Academy of Sciences
- 2019–2026University of South China
- 2009–2011Hong Kong Polytechnic UniversityPostdoctoral fellow
Education
- 2008Harbin Institute of TechnologyPhD
- 2002Harbin Institute of TechnologyMaster's degree
- 1999Harbin Institute of TechnologyBachelor's degree
Most-cited works
- ECA-Net: Efficient Channel Attention for Deep Convolutional Neural NetworksWith Wangmeng Zuo
- Enhancing Geometric Factors in Model Learning and Inference for Object Detection and Instance SegmentationWith Wangmeng Zuo
- Detection and Tracking Meet Drones ChallengeWith Haibin Ling
- Neighborhood rough set based heterogeneous feature subset selection
- Progressive Image Deraining Networks: A Better and Simpler BaselineWith Deyu Meng and Wangmeng Zuo
Colleagues
Sources and credits
Photo: We have no published photo of Qinghua Hu yet. Send us a free-licensed photo.