Adapter Merging with Centroid Prototype Mapping for Scalable Class-Incremental Learning
Takuma Fukuda, Hiroshi Kera, Kazuhiko Kawamoto
Abstract
We propose Adapter Merging with Centroid Prototype Mapping (ACMap), an exemplar-free framework for classincremental learning (CIL) that addresses both catastrophic forgetting and scalability. While existing methods involve a trade-off between inference time and accuracy, ACMap consolidates task-specific adapters into a single adapter, thus achieving constant inference time across tasks without sacrificing accuracy. The framework employs adapter merging to build a shared subspace that aligns task representations and mitigates forgetting, while centroid prototype mapping maintains high accuracy by consistently adapting representations within the shared subspace. To further improve scalability, an early stopping strategy limits adapter merging as tasks increase. Extensive experiments on five benchmark datasets demonstrate that ACMap matches state-ofthe-art accuracy while maintaining inference time comparable to the fastest existing methods. The code is available at https://github.com/tf63/ACMap . To address the dual challenges of catastrophic forgetting and scalability in CIL, we propose Adapter Merging with Centroid Prototype Mapping (ACMap), a framework that consolidates task-specific adapters into a single adapter. ACMap enables scalability by maintaining con-
Ask Lune
Know this paper in minutes, not an afternoon.
Lune can read every page of it and the top-tier papers that cite it, and gives you its claims, its evidence and its limits, each one cited to the paper it comes from. No skimming, no AI slop.
10 top-tier papers have cited it since. Know what they found before your reviewers do.
Lune can draw on
the full paper · the top-tier papers that cite it · related work
Free to start. No credit card required.
Lineage
- SECOS: Semantic Capture for Rigorous Classification in Open-World Semi-Supervised Learning cites An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
- Multi-Cache Enhanced Prototype Learning for Test-Time Generalization of Vision-Language Models cites The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalization
- Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time cites An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
- Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference time cites The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution Generalization
- MINGLE: Mixture of Null-Space Gated Low-Rank Experts for Test-Time Continual Model Merging cites LoRA: Low-Rank Adaptation of Large Language Models
- AdaptFormer: Adapting Vision Transformers for Scalable Visual Recognition cites An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
- Harnessing Textual Semantic Priors for Knowledge Transfer and Refinement in CLIP-Driven Continual Learning cites An Image is Worth 16x16 Words: Transformers for Image Recognition at Scale
- Harnessing Textual Semantic Priors for Knowledge Transfer and Refinement in CLIP-Driven Continual Learning cites AdaptFormer: Adapting Vision Transformers for Scalable Visual Recognition
Cited by top-tier papers10
- MINGLE: Mixture of Null-Space Gated Low-Rank Experts for Test-Time Continual Model MergingZihuan Qiu, Yi Xu, Chiyuan He, Fanman Meng et al.NeurIPS 2025 · 16 citations
- Looking Beyond the Known: Towards a Data Discovery Guided Open-World Object DetectionAnay Majee, Amitesh Gangrade, Rishabh IyerNeurIPS 2025 · 5 citations
- Multi-Cache Enhanced Prototype Learning for Test-Time Generalization of Vision-Language ModelsXinyu Chen, Haotian Zhai, Can Zhang, Xiupeng Shi et al.ICCV 2025 · 2 citations
- Harnessing Textual Semantic Priors for Knowledge Transfer and Refinement in CLIP-Driven Continual LearningLingfeng He, De Cheng, Di Xu, Huaijie Wang et al.AAAI 2026 · 1 citation
- SECOS: Semantic Capture for Rigorous Classification in Open-World Semi-Supervised LearningHezhao Liu, Jiacheng Yang, Junlong Gao, Mengke Li et al.CVPR 2026
Builds on25
- An Image is Worth 16x16 Words: Transformers for Image Recognition at ScaleAlexey Dosovitskiy, Lucas Beyer, Alexander Kolesnikov, Dirk Weissenborn et al.ICLR 2021 · 21,477 citations
- LoRA: Low-Rank Adaptation of Large Language ModelsEdward J. Hu, Yelong Shen, Phillip Wallis, Zeyuan Allen-Zhu et al.ICLR 2022 · 18,833 citations
- The Many Faces of Robustness: A Critical Analysis of Out-of-Distribution GeneralizationDan Hendrycks, Steven Basart, Norman Mu, Saurav Kadavath et al.ICCV 2021 · 2,294 citations
- Model soups: averaging weights of multiple fine-tuned models improves accuracy without increasing inference timeMitchell Wortsman, Gabriel Ilharco, Samir Yitzhak Gadre, Rebecca Roelofs et al.ICML 2022 · 1,464 citations
- AdaptFormer: Adapting Vision Transformers for Scalable Visual RecognitionShoufa Chen, Chongjian Ge, Zhan Tong, Jiangliu Wang et al.NeurIPS 2022 · 1,291 citations
Related papers
- Exemplar-Free Class Incremental Learning via Preserving Class-Discriminative StructureXin Zhang, Liang Bai, Guanchao Wang, Xian YangCVPR 2026
- Representation-Steered Incremental Adapter-Tuning for Class-Incremental Learning with Pre-Trained ModelsJiarui Zhao, Libo Huang, Xiangqi Li, Zhulin An et al.CVPR 2026 · 1 citation
- Bring Evanescent Representations to Life in Lifelong Class Incremental LearningMarco Toldo, Mete OzayCVPR 2022 · 34 citations
- GACL: Exemplar-Free Generalized Analytic Continual LearningHuiping Zhuang, Yizhu Chen, Di Fang, Run He et al.NeurIPS 2024 · 36 citations
- AnaCP: Toward Upper-Bound Continual Learning via Analytic Contrastive ProjectionSaleh Momeni, Changnan Xiao, Bing LiuNeurIPS 2025 · 8 citations
