Research interets

Machine Learning

Machine learning embodies a remarkably elegant theory. My research interests primarily focus on unraveling the operating principles behind neural networks in deep learning. My most recent lines of inquiry involve Information Geometry and Mean Field Theory, both of which have been deeply inspiring to me.

Computer Vision

Computer vision is currently profoundly influenced by advancements in deep learning. My work centers on exploring how to synergize deep learning and computer vision to efficiently and robustly accomplish tasks such as image fusion, image classification, and semantic segmentation.

Semantic Communication

Semantic communication is an emerging cross-generational communication technology. With the evolution of deep learning techniques, it has become a consensus to view the weights of neural networks as shared knowledge bases. This allows communication systems to leverage the advantages of neural networks to break through the Shannon limit. My work is focused on designing more efficient semantic communication systems and contributing to the refinement of the fundamental theoretical framework for semantic communication.