Hi! I am a first-year PhD student at University of Electronic Science and Technology of China (UESTC). I am advised by Hongliang Li. Previously, I received my B.E in Electronic Information Engineering at UESTC.
My research interests include continual learning, transfer learning, their applications to multimodality and object detection, and social network analysis. Currently, my research focuses on continual learning. Specifically, I want to understand underlying mechanisms behind catastrophic forgetting of deep models in learning knowledge continuously, and design methods and systems to tackle such catastrophic forgetting problem.
PhD in Machine Learning, 2022 - Present
University of Electronic Science and Technology of China
MSc in Information and Communication, 2020 - 2022
University of Electronic Science and Technology of China
BEng in Electronic Information Engineering, 2016 - 2020
University of Electronic Science and Technology of China
This work is the first to consider a new type of bias - task-induced bias in a causal perspective, which is derived from special settings of continual learning. We analyzed three continual scenarios in a causal framework, and found that the task-induced bias is particularly detrimental to CIL. Based on this framework, we designed a causal intervention operation and implemented it as a causal debias module by exploiting a powerful attention mechanism.