Research map / Label-efficient and robust learning / Test-time adaptation
AdaNPC: Exploring Non-Parametric Classifier for Test-Time Adaptation
YiFan Zhang, xue wang, Kexin Jin, Kun Yuan, Zhang Zhang, Liang Wang, Rong Jin, Tieniu Tan
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 |
|---|---|---|---|
| Replace the linear classifier with AdaNPC KNN | ResNet50 test-time adaptation · PACS, BSZ=32 | accuracy: 86.68 -> 88.93 | Improved |
| Use AdaNPC rather than the strongest listed DG baseline | domain generalization models · OOD benchmarks average | accuracy: 71.7 -> 72.7 | Improved |
| Add BN retraining at test time (ablation) | AdaNPC · OOD benchmarks average | accuracy: 72.7 -> 72.9 | Improved |
| Add BN retraining at test time (ablation) | AdaNPC · PACS, ResNet50 | accuracy: 85.8 -> 85.7 | No clear effect |
| Train representations with KNN loss rather than ERM cross-entropy (ablation) | AdaNPC · PACS test-domain validation | accuracy: 84.2 -> 86.1 | Improved |
| Vary the KNN neighbor count and memory size (ablation) | KNN-loss AdaNPC · PACS test-domain validation | accuracy: 85.2; 85.1; 86.1; 84.5; 85.7 | Mixed |
| Use AdaNPC instead of gradient-updated full-model pseudo-label adaptation | ResNet50 TTA · PACS, BSZ=2 | accuracy: 73.35 -> 88.93 | Improved |
| Add online target features and predicted labels to memory | AdaNPC · successive adaptation on Rotated MNIST | accuracy: +50% | Improved |
Closest papers by meaning
- DomainAdaptor: A Novel Approach to Test-time Adaptation (ICCV 2023 · similarity 0.89)
- Test-Time Domain Adaptation by Learning Domain-Aware Batch Normalization (AAAI 2024 · similarity 0.87)
- A Layer Selection Approach to Test Time Adaptation (AAAI 2025 · similarity 0.87)
- Efficient Test-time Adaptation through Candidate Verification and Divergence Shifts (NeurIPS 2026 · similarity 0.86)
- Improved Self-Training for Test-Time Adaptation (CVPR 2024 · similarity 0.86)
- Protected Test-Time Adaptation via Online Entropy Matching: A Betting Approach (NeurIPS 2024 · similarity 0.86)
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.
