
Zhongyi Han
out-of-distribution detection
source-free domain adaptation
discriminability and transferability estimation
a bayesian perspective
channel-aware typical features
test-time detection
3
presentations
SHORT BIO
Zhongyi Han received the B.Eng. and M.M. degrees from the Shandong University of Traditional Chinese Medicine, in 2016 and 2019, respectively. He is currently pursuing the Ph.D. degree with the Artificial Intelligence Research Center, Shandong University, supervised by Prof. Yilong Yin. He had been a Visiting Student at the University of Western Ontario, Canada. He has published over twenty papers in leading international journals and conferences. His research interests include machine learning, computer vision, and medical image analysis.
Presentations

Discriminability and Transferability Estimation: A Bayesian Source Importance Estimation Approach for Multi-Source-Free Domain Adaptation
Zhongyi Han and 6 other authors

Not All Parameters Should be Treated Equally: Deep Safe Semi-Supervised Learning under Class Distribution Mismatch
Rundong He and 3 other authors

Exploring Channel-Aware Typical Features for Out-of-Distribution Detection
Rundong He and 7 other authors