Research map / Label-efficient and robust learning / Recognition and long-tail
Learning Like Humans: Analogical Concept Learning for Generalized Category Discovery
Jizhou Han, Chenhao Ding, Yuhang He, Qiang Wang, Shaokun Wang, SongLin Dong, Yihong Gong
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 |
|---|---|---|---|
| attach AL-GCD (ATCG) | SelEx-CLIP · CUB | accuracy (All): 74.2 -> 84.1 | Improved |
| attach AL-GCD (ATCG) | SimGCD-CLIP · Stanford Cars | accuracy (All): 69.4 -> 78.3 | Improved |
| attach AL-GCD (ATCG) | CMS-CLIP · FGVC Aircraft | accuracy (All): 50.3 -> 54.8 | Improved |
| attach AL-GCD (ATCG) | SelEx-CLIP · CIFAR100 | accuracy (All): 77.2 -> 80.4 | Improved |
| attach AL-GCD (ATCG) | SelEx-CLIP · CIFAR100 | novel accuracy: 77.6 -> 69.8 | Hurt |
| attach AL-GCD across all pipelines | CLIP-based GCD pipelines · All-datasets average (K known) | accuracy (All): avg +5.0 | Improved |
| attach AL-GCD | CMS-CLIP · All datasets (K unknown) | accuracy (All): 65.8 -> 69.8 | Improved |
| add ATCG initial layer (ablation) | SimGCD-CLIP · CIFAR100 | novel accuracy: 75.2 -> 82.7 | Improved |
| add ATCG stacked layers on top of initial layer (ablation) | SimGCD-CLIP + ATCG · ImageNet100 | accuracy (All): 90.4 -> 92.6 | Improved |
| add ATCG stacked layers on top of initial layer (ablation) | SimGCD-CLIP + ATCG · Stanford Cars | accuracy (All): 74.9 -> 78.3 | Improved |
| increase ATCG depth from 2 to 4/6 layers (ablation) | SimGCD-CLIP + ATCG · CUB | accuracy (All): 74.73 -> 74.80 (4) / 74.69 (6) | No clear effect |
| increase vision weight alpha from 0.4 to 0.7 (ablation) | AL-GCD · Stanford Cars | old / new accuracy: old +1.8, new -1.7 | Mixed |
Closest papers by meaning
- Active Generalized Category Discovery (CVPR 2024 · similarity 0.89)
- ProtoGCD: Unified and Unbiased Prototype Learning for Generalized Category Discovery (TPAMI 2025 · similarity 0.89)
- Transfer and Alignment Network for Generalized Category Discovery (AAAI 2024 · similarity 0.89)
- Solving the Catastrophic Forgetting Problem in Generalized Category Discovery (CVPR 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)
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.
