Research map / Label-efficient and robust learning / Recognition and long-tail
Bures Generalized Category Discovery
Luyao Tang, Kunze Huang, Chaoqi Chen, Cheng Chen
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add BIA loss | SimGCD · CUB (K given) | All accuracy (%): 60.7 -> 62.1 | Improved |
| add BIA loss | CMS · CUB (K given) | All accuracy (%): 67.1 -> 71.1 | Improved |
| add BIA loss | SelEx · CUB (K given) | All accuracy (%): 78.7 -> 80.6 | Improved |
| add BIA loss | SPTNet · FGVC Aircraft (K given) | All accuracy (%): 51.6 -> 54.7 | Improved |
| add BIA loss | SimGCD · ImageNet100 (K given) | All accuracy (%): 83.3 -> 86.7 | Improved |
| add BIA loss | SimGCD · CIFAR100 (K given) | All accuracy (%): 80.1 -> 80.2 | No clear effect |
| add BIA loss | CMS · CIFAR100 (K given) | All accuracy (%): 79.5 -> 79.0 | Hurt |
| add BIA loss | CMS · Herbarium 19 (K given) | All accuracy (%): 36.5 -> 36.3 | No clear effect |
| add BIA loss on average over fine-grained datasets | GCD baselines · CUB (K given) | average change in All accuracy: +2.2 | Improved |
| add BIA loss to estimate number of classes | CMS · ImageNet100 K estimation | error rate (%): 2 -> 0 | Improved |
| add BIA loss to estimate number of classes | CMS · Stanford Cars K estimation | error rate (%): 23.9 -> 18.9 | Improved |
| use BIA instead of VICReg as isotropy regularizer (ablation) | CMS · CUB (isotropy comparison) | All accuracy (%): 68.3 -> 71.1 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- SPTNet — SPTNet: An Efficient Alternative Framework for Generalized Category Discovery with Spatial Prompt Tuning (ICLR 2024) · better on 4 of 4 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Hyperbolic Category Discovery (CVPR 2025 · similarity 0.89)
- Towards Understanding Parametric Generalized Category Discovery on Graphs (ICML 2025 · similarity 0.88)
- Happy: A Debiased Learning Framework for Continual Generalized Category Discovery (NeurIPS 2024 · similarity 0.88)
- Continual Generalized Category Discovery: Learning and Forgetting from a Bayesian Perspective (ICML 2025 · similarity 0.88)
- Consistent Supervised-Unsupervised Alignment for Generalized Category Discovery (NeurIPS 2025 · similarity 0.87)
- Generalized Category Discovery under Domain Shift: A Frequency Domain Perspective (NeurIPS 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
