Reproduce “Semantic Mismapping in Generalized Category Discovery: Evaluation and Mitigation” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Zitai Wang, Qianqian Xu, Cong Hua, Xi Chen, Zhiyong Yang, Xiaochun Cao, Qingming Huang
The authors propose Semantics-Preserving Evaluation (SPE), which keeps supervised old-category outputs fixed and matches only novel outputs, and Mismapping-Mitigating Optimization (MMO), which adds losses aimed at mismapping. On Herbarium 19, SimGCD with a CLIP backbone’s new-category accuracy is 38.26% under global Hungarian matching and 12.63% under SPE; MMO raises SPE accuracy to 43.56%. Gaps vary by dataset, and gains are not universal.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.