Research map / Continual learning / Class-incremental
Consistency-Driven Calibration and Matching for Few-Shot Class Incremental Learning
Qinzhe Wang, Zixuan Chen, Keke Huang, Xiu Su, Chunhua Yang, Chang Xu
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 memory-aware prototype calibration (MPC) (ablation) | ConCM (projector only) · mini-ImageNet | AHM: 47.83 -> 52.35 | Improved |
| add dynamic structure matching (DSM) (ablation) | ConCM (projector only) · mini-ImageNet | AHM: 47.83 -> 56.79 | Improved |
| add both MPC and DSM (ablation) | ConCM (projector only) · mini-ImageNet | AHM: 47.83 -> 59.78 | Improved |
| add MPC on top of DSM (ablation) | ConCM · CIFAR100 | AHM: 56.96 -> 59.05 | Improved |
| add MPC on top of DSM (ablation) | ConCM · CUB200 | AHM: 60.19 -> 62.20 | Improved |
| add fine-tuned projector to frozen backbone (ablation) | ConCM baseline · mini-ImageNet | AHM: 22.00 -> 47.83 | Improved |
| replace cross-entropy with matching loss (ablation) | ConCM · mini-ImageNet | AHM: 47.83 -> 55.28 | Improved |
| add contrastive loss with anchor to matching loss (ablation) | ConCM · mini-ImageNet | AHM: 55.28 -> 59.78 | Improved |
| use anchor contrastive loss instead of general SCL loss (ablation) | ConCM · mini-ImageNet | AHM: 59.42 -> 59.78 | Improved |
| use DSM only when class names absent from WordNet (no MPC) | ConCM w/o MPC vs NC-FSCIL · mini-ImageNet | AHM: 52.62 -> 56.79 | Improved |
| ConCM versus best baseline OrCo | ConCM · CIFAR100 | final-session accuracy: 57.51 -> 58.33 | Improved |
| ConCM versus OrCo on first incremental session | ConCM · CIFAR100 | session-1 harmonic mean: 72.11 -> 72.27 | No clear effect |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- Mamba-FSCIL — Mamba-FSCIL: Dynamic Adaptation With Selective State Space Model for Few-Shot Class Incremental Learning (TPAMI 2026) · better on 4 of 4 shared results
- OrCo — OrCo: Towards Better Generalization via Orthogonality and Contrast for Few-Shot Class-Incremental Learning (CVPR 2024) · better on 4 of 4 shared results
- CLOSER — CLOSER: Towards Better Representation Learning for Few-Shot Class-Incremental Learning (ECCV 2024) · better on 2 of 2 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Few-Shot Class-Incremental Learning via Training-Free Prototype Calibration (NeurIPS 2023 · similarity 0.90)
- Navigating Semantic Drift in Task-Agnostic Class-Incremental Learning (ICML 2025 · similarity 0.90)
- HyCal: A Training-Free Prototype Calibration Method for Cross-Discipline Few-Shot Class-Incremental Learning (CVPR 2026 · similarity 0.90)
- Advancing Analytic Class-Incremental Learning through Vision-Language Calibration (ICML 2026 · similarity 0.89)
- Tripartite Weight-Space Ensemble for Few-Shot Class-Incremental Learning (CVPR 2025 · similarity 0.89)
- Enhancing Few-Shot Class-Incremental Learning via Training-Free Bi-Level Modality Calibration (CVPR 2025 · similarity 0.88)
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.
