Research map / Continual learning / Class-incremental
Self-Promoted Prototype Refinement for Few-Shot Class-Incremental Learning
Kai Zhu, Yang Cao, Wei Zhai, Jie Cheng, Zheng-Jun Zha
Check 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 |
|---|---|---|---|
| Use self-promoted prototype refinement instead of fine-tuning for prototype updates, with RESS (ablation) | Incremental prototype learning · CIFAR-100 | average accuracy: 51.00 -> 54.51 | Improved |
| Add RESS to standard representation with SPPR (ablation) | Incremental prototype learning · CIFAR-100 | average accuracy: 50.20 -> 54.44 | Improved |
| Add SPPR to RESS with fine-tuning (ablation) | Incremental prototype learning · CIFAR-100 | average accuracy: 51.00 -> 54.51 | Improved |
| Add fine-tuning to RESS + SPPR (ablation) | Incremental prototype learning · CIFAR-100 | average accuracy: 54.44 -> 54.51 | No clear effect |
| Increase the number of prototype-update iterations (ablation) | RESS feature representation · CIFAR-100 | accuracy | Hurt |
| Use 1-shot rather than more-shot episodic training (ablation) | RESS · CIFAR-100 | accuracy | Hurt |
| Increase episodic training ways beyond 5 (ablation) | RESS · CIFAR-100 | accuracy | Hurt |
| Increase incremental test shots from 1 to 5 or 10 (ablation) | Prototype-based inference · CIFAR-100 | accuracy | No clear effect |
| Use dynamic prototype refinement rather than zero, random, or unchanged prototypes (ablation) | Prototype update · CIFAR-100 | accuracy | Improved |
Closest papers by meaning
- DiAPR: Dimensionally-Allocated Prototype Refinement for Non-Exemplar Class Incremental Learning (AAAI 2026 · similarity 0.90)
- Pseudo Informative Episode Construction for Few-Shot Class-Incremental Learning (AAAI 2025 · similarity 0.90)
- Self-Sustaining Representation Expansion for Non-Exemplar Class-Incremental Learning (CVPR 2022 · similarity 0.90)
- Non-exemplar Online Class-Incremental Continual Learning via Dual-Prototype Self-Augment and Refinement (AAAI 2024 · similarity 0.87)
- M2SD:Multiple Mixing Self-Distillation for Few-Shot Class-Incremental Learning (AAAI 2024 · similarity 0.87)
- Tripartite Weight-Space Ensemble for Few-Shot Class-Incremental Learning (CVPR 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.
