Akang Wang

Akang Wang

Yunnan Normal University

Hi everyone, Iā€™m currently working on 3D scene reconstruction and understanding. My prior research covers open-vocabulary segmentation and multi-label learning, and now Iā€™m applying my 2D semantic understanding expertise to 3D scene research.

šŸ”„ News

  • 2026.05: Ā šŸŽ‰šŸŽ‰ PIAA has been accepted by the International Conference on Machine Learning (ICML).

šŸ“ Publications

ICML 2026
sym

[CLS] is Not Enough: Multi-Label Recognition via Patch-Level Inference and Adaptive Aggregation

Akang Wang, Xili Deng, Zhanxuan Hu, YiZhao, Yonghang Tai, Huafeng Li

Project

  • We tackle multi-label recognition by analytically deriving a patch-based visual classifier to generate reliable, fine-grained object predictions without any model training. We then adaptively fuse these local patch scores with the global image context to robustly identify multiple co-existing targets.

šŸ“– Educations

  • 2025.09 - now, Master of Computer Science and Technology Yunnan Normal University, Kunming, China.