Chenxun Deng

Ph.D. Student, Institute of Automation, Chinese Academy of Sciences

I work on computer vision and machine learning, with a focus on long-tailed recognition, diffusion models, wildlife visual understanding, and multi-animal tracking. My recent research explores how prior knowledge, generative modeling, and expert specialization can improve recognition under imbalanced data and complex real-world conditions.

Research Interests

Computer Vision Long-Tailed Learning Diffusion Models Wildlife Recognition Multi-Animal Tracking Vision-Language Modeling

About

My research centers on robust visual understanding under challenging data conditions, especially long-tailed distributions, limited supervision, and complex biological or ecological scenes. I am also interested in the interaction between generative modeling and recognition, particularly how diffusion models and language-guided priors can support discriminative tasks.

Selected Publications