Yushan Lai

Education

2021-2025 (current)

BSc, Artificial Intelligence; Sun Yat-sen University

• Advisor: Prof.Juepeng Zheng

2025-2030 (expected)

PH.D, Computer Science and Technology; Tsinghua University

• Advisor: Prof.Haohuan Fu

Research Experience

ADU: Adaptive Detection of Unknown Categories in Black-Box Domain Adaptation

• This research proposes ADU, a novel framework for black-box domain adaptation that adaptively detects unknown target categories, achieving SOTA performance while preserving source data privacy.
• Advisor: Prof.Juepeng Zheng

C3DA: A Universal Domain Adaptation Method for Scene Classification From Remote Sensing Imagery

• This study introduces C3DA, a universal domain adaptation method for remote sensing scene classification that improves recognition of unknown classes by integrating confidence, consistency, and certainty, achieving superior performance across multiple datasets and tasks.
• Advisor: Prof.Juepeng Zheng

Publication

[1] Yushan Lai, Guowen Li, Haoyuan Liang, Juepeng Zheng, and Zhiyu Ye. Adu: Adaptive detection of unknown categories in black-box domain adaptation. In Proceedings of the Computer Vision and Pattern Recognition Conference (CVPR), pages 30588–30598, 2025.

[2] Jiaxu Guo, Yushan Lai, Jinxiao Zhang, Juepeng Zheng, Haohuan Fu, Lin Gan, Liang Hu, Gaochao Xu, and Xilong Che. C 3 da: A universal domain adaptation method for scene classification from remote sensing imagery. IEEE Geoscience and Remote Sensing Letters, 2024.