Research map / Label-efficient and robust learning / Recognition and long-tail
This Looks Distinctly Like That: Grounding Interpretable Recognition in Stiefel Geometry against Neural Collapse
Junhao Jia, Jiaqi Wang, Yunyou Liu, Haodong Jing, Yueyi Wu, Xian Wu, Yefeng Zheng
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 Stiefel manifold constraint on prototypes (ablation) | AMP · CUB-200-2011 | accuracy: 85.2 -> 88.4 | Improved |
| add Stiefel manifold constraint on prototypes (ablation) | AMP · CUB-200-2011 | Consistency: 65.0 -> 76.8 | Improved |
| add Stiefel manifold constraint on prototypes (ablation) | AMP · CUB-200-2011 | DAUC: 4.95 -> 3.45 | Improved |
| add Stiefel manifold constraint on prototypes (ablation) | AMP · Stanford Cars | accuracy: 90.2 -> 92.0 | Improved |
| add dynamic capacity matrix Sigma_c with rank calibration (ablation) | AMP · Stanford Cars | OIRR: 35.4 -> 33.5 | Improved |
| add dynamic capacity matrix Sigma_c with rank calibration (ablation) | AMP · CUB-200-2011 | accuracy: 87.9 -> 88.4 | Improved |
| add spatial entropy minimization (ablation) | AMP · CUB-200-2011 | accuracy: 87.0 -> 88.4 | Improved |
| add spatial entropy minimization (ablation) | AMP · CUB-200-2011 | DAUC: 3.92 -> 3.45 | Improved |
| add spatial overlap penalty (ablation) | AMP · CUB-200-2011 | accuracy: 88.1 -> 88.4 | Improved |
| add spatial overlap penalty (ablation) | AMP · Stanford Cars | OIRR: 34.0 -> 33.5 | Improved |
| add spatial overlap penalty (ablation) | AMP · Stanford Cars | accuracy: 91.6 -> 92.0 | Improved |
Closest papers by meaning
- Engineering the Neural Collapse Geometry of Supervised-Contrastive Loss (Student Abstract) (AAAI 2024 · similarity 0.87)
- Guiding Neural Collapse: Optimising Towards the Nearest Simplex Equiangular Tight Frame (NeurIPS 2024 · similarity 0.86)
- Neural Collapse in Cumulative Link Models for Ordinal Regression: An Analysis with Unconstrained Feature Model (NeurIPS 2025 · similarity 0.85)
- Generalized Neural Collapse for a Large Number of Classes (ICML 2024 · similarity 0.85)
- Neural Collapse Beyond the Unconstrained Features Model: Landscape, Dynamics, and Generalization in the Mean-Field Regime (ICML 2025 · similarity 0.85)
- The Geometric Wall: Manifold Structure Predicts Layerwise Sparse Autoencoder Scaling Laws (NeurIPS 2026 · similarity 0.85)
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.
