Research map / Continual learning / Class-incremental
Learning With Fantasy: Semantic-Aware Virtual Contrastive Constraint for Few-Shot Class-Incremental Learning
Read the paperCheck 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 |
|---|---|---|---|
| Add supervised contrastive learning to CE training (ablation) | FSCIL model · miniImageNet | final-session accuracy (%): 45.54 -> 47.10 | Helped |
| Add semantic-aware virtual-class fantasy to CE+SCL (ablation) | FSCIL model · miniImageNet | final-session accuracy (%): 47.10 -> 54.52 | Helped |
| Add multi-crop training (ablation) | SAVC · miniImageNet | final-session accuracy (%): 54.52 -> 55.80 | Helped |
| Add incremental-session finetuning (ablation) | SAVC · miniImageNet | final-session accuracy (%): 55.80 -> 57.11 | Helped |
| Use more complex transformations for virtual classes (ablation) | SAVC · miniImageNet | class separation | Helped |
| Use more complex transformations for virtual classes (ablation) | SAVC · CUB200 and CIFAR100 | class separation | Hurt |
| Finetune fewer and deeper layers during incremental sessions instead of too many layers (ablation) | SAVC · miniImageNet | accuracy | Helped |
| Finetune too many layers during incremental sessions (ablation) | SAVC · miniImageNet | accuracy | Hurt |
| Replace the CE baseline with SAVC | FSCIL model · CUB200, final session | accuracy (%): 56.94 -> 62.50 | Helped |
| Use semantic-aware virtual classes and fantasy-space inference | FSCIL model · CIFAR100, CUB200, and miniImageNet | final-session accuracy: 53.12%; 62.50%; 57.11% | Helped |
Closest papers by meaning
- OrCo: Towards Better Generalization via Orthogonality and Contrast for Few-Shot Class-Incremental Learning (CVPR 2024 · similarity 0.89)
- BECLR: Batch Enhanced Contrastive Few-Shot Learning (ICLR 2024 · similarity 0.88)
- Self-Supervised Contrastive Learning is Approximately Supervised Contrastive Learning (NeurIPS 2025 · similarity 0.88)
- Envisioning Class Entity Reasoning by Large Language Models for Few-shot Learning (AAAI 2025 · similarity 0.87)
- Pre-trained Vision and Language Transformers Are Few-Shot Incremental Learners (CVPR 2024 · similarity 0.87)
- Tripartite Weight-Space Ensemble for Few-Shot Class-Incremental Learning (CVPR 2025 · 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.