Research map / Continual learning / Class-incremental
Few-Shot Class-Incremental Learning
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| represent old knowledge with neural gas topology and stabilize it using anchor loss | TOPIC · CIFAR100, final session, QuickNet | accuracy: 24.17% vs 19.50% | Helped |
| represent old knowledge with neural gas topology and stabilize it using anchor loss | TOPIC · CIFAR100, final session, ResNet18 | accuracy: 29.37% vs 15.85% | Helped |
| represent old knowledge with neural gas topology and stabilize it using anchor loss | TOPIC · miniImageNet, final session, QuickNet | accuracy: 18.36% vs 13.59% | Helped |
| represent old knowledge with neural gas topology and stabilize it using anchor loss | TOPIC · miniImageNet, final session, ResNet18 | accuracy: 24.42% vs 19.58% | Helped |
| represent old knowledge with neural gas topology and stabilize it using anchor loss | TOPIC · CUB200, session 11, ResNet18 | accuracy: 26.28 vs 22.11 | Helped |
| add the MML min and max terms (ablation) | anchor-loss TOPIC · miniImageNet, final session, ResNet18 | accuracy: 21.80 -> 24.42 | Helped |
| add the MML min term (ablation) | anchor-loss TOPIC · miniImageNet, final session, ResNet18 | accuracy: 21.80 -> 24.18 | Helped |
| add the MML max term (ablation) | anchor-loss TOPIC · miniImageNet, final session, ResNet18 | accuracy: 21.80 -> 23.80 | Helped |
| remove the variance re-weighting matrix Λ from anchor loss (ablation) | TOPIC · miniImageNet, final session, ResNet18 | accuracy: 21.80 -> 21.36 | Hurt |
| add distillation loss to AL-MML (ablation) | TOPIC · miniImageNet, final session, ResNet18 | accuracy: 24.42 -> 23.49 | Hurt |
| use neural gas nodes instead of exemplars for knowledge representation (ablation) | TOPIC · CIFAR100, final session, ResNet18, memory size 400 | accuracy: 28.25 -> 29.37 | Helped |
| add anchor loss to naive finetuning (ablation) | TOPIC · FSCIL across datasets and backbones | accuracy: up to 38.90% | Helped |
Closest papers by meaning
- Few-Shot Class-Incremental Learning via Class-Aware Bilateral Distillation (CVPR 2023 · similarity 0.93)
- Tripartite Weight-Space Ensemble for Few-Shot Class-Incremental Learning (CVPR 2025 · similarity 0.90)
- Pre-trained Vision and Language Transformers Are Few-Shot Incremental Learners (CVPR 2024 · similarity 0.88)
- M2SD:Multiple Mixing Self-Distillation for Few-Shot Class-Incremental Learning (AAAI 2024 · similarity 0.88)
- Adaptive Decision Boundary for Few-Shot Class-Incremental Learning (AAAI 2025 · similarity 0.88)
- Pseudo Informative Episode Construction for Few-Shot Class-Incremental Learning (AAAI 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.