Research map / Continual learning / Class-incremental
Few-shot Class Incremental Learning with Attention-Aware Self-Adaptive Prompt
Chenxi Liu, Zhenyi Wang, Tianyi Xiong, Ruibo Chen, Yihan Wu, junfeng guo, Heng Huang
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 attention-aware task-invariant prompt (TIP) (ablation) | ASP · ImageNet-R FSCIL | A_avg (%): 72.2 -> 75.3 | Improved |
| add self-adaptive task-specific prompt (TSP) (ablation) | ASP · CIFAR100 FSCIL | A_avg (%): 88.1 -> 89.0 | Improved |
| add anchor loss L_c (ablation) | ASP · ImageNet-R FSCIL | A_avg (%): 74.5 -> 75.3 | Improved |
| initialize all TIP tokens with the same value instead of random different values (ablation) | ASP · CIFAR100 FSCIL | A_avg (%): 87.6 -> 89.0 | Improved |
| replace FC classifier head with class-mean prototypes in prompt-based CIL baselines | DualP · CIFAR100 FSCIL | HAcc (%): 0.1 -> 75.3 | Improved |
| replace FC classifier head with class-mean prototypes in prompt-based CIL baselines | L2P · ImageNet-R FSCIL | HAcc (%): 1.0 -> 52.0 | Improved |
| increase incremental shots from 5 to 20 (ablation) | ASP · ImageNet-R FSCIL | A_avg (%): 75.3 -> 77.4 | Improved |
| increase incremental shots from 1 to 5 (ablation) | ASP · CUB200 FSCIL | A_avg (%): 78.6 -> 83.8 | Improved |
| use ASP prompts instead of DualP+ on ImageNet-R | ASP vs DualP+ · ImageNet-R FSCIL | PD: 9.0 vs 13.5 | Hurt |
| keep average prompt features p_avg and EMA update for new tasks (ablation) | ASP · ImageNet-R FSCIL | A_avg, PD, HAcc | Improved |
Closest papers by meaning
- SEC-Prompt:SEmantic Complementary Prompting for Few-Shot Class-Incremental Learning (CVPR 2025 · similarity 0.91)
- Pre-trained Vision and Language Transformers Are Few-Shot Incremental Learners (CVPR 2024 · similarity 0.90)
- PEARL: Input-Agnostic Prompt Enhancement with Negative Feedback Regulation for Class-Incremental Learning (AAAI 2025 · similarity 0.88)
- Tripartite Weight-Space Ensemble for Few-Shot Class-Incremental Learning (CVPR 2025 · similarity 0.87)
- Adaptive Decision Boundary for Few-Shot Class-Incremental Learning (AAAI 2025 · similarity 0.87)
- CAPrompt: Cyclic Prompt Aggregation for Pre-Trained Model Based 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.
