MoTiC: Momentum Tightness and Contrast for Few-Shot Class-Incremental Learning
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866911417051906048 |
|---|---|
| author | He, Zeyu Huang, Shuai Lu, Yuwu Zhao, Ming |
| author_facet | He, Zeyu Huang, Shuai Lu, Yuwu Zhao, Ming |
| contents | Few-Shot Class-Incremental Learning (FSCIL) must contend with the dual challenge of learning new classes from scarce samples while preserving old class knowledge. Existing methods use the frozen feature extractor and class-averaged prototypes to mitigate against catastrophic forgetting and overfitting. However, new-class prototypes suffer significant estimation bias due to extreme data scarcity, whereas base-class prototypes benefit from sufficient data. In this work, we theoretically demonstrate that aligning the new-class priors with old-class statistics via Bayesian analysis reduces variance and improves prototype accuracy. Furthermore, we propose large-scale contrastive learning to enforce cross-category feature tightness. To further enrich feature diversity and inject prior information for new-class prototypes, we integrate momentum self-supervision and virtual categories into the Momentum Tightness and Contrast framework (MoTiC), constructing a feature space with rich representations and enhanced interclass cohesion. Experiments on three FSCIL benchmarks produce state-of-the-art performances, particularly on the fine-grained task CUB-200, validating our method's ability to reduce estimation bias and improve incremental learning robustness. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_19664 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | MoTiC: Momentum Tightness and Contrast for Few-Shot Class-Incremental Learning He, Zeyu Huang, Shuai Lu, Yuwu Zhao, Ming Computer Vision and Pattern Recognition Artificial Intelligence Few-Shot Class-Incremental Learning (FSCIL) must contend with the dual challenge of learning new classes from scarce samples while preserving old class knowledge. Existing methods use the frozen feature extractor and class-averaged prototypes to mitigate against catastrophic forgetting and overfitting. However, new-class prototypes suffer significant estimation bias due to extreme data scarcity, whereas base-class prototypes benefit from sufficient data. In this work, we theoretically demonstrate that aligning the new-class priors with old-class statistics via Bayesian analysis reduces variance and improves prototype accuracy. Furthermore, we propose large-scale contrastive learning to enforce cross-category feature tightness. To further enrich feature diversity and inject prior information for new-class prototypes, we integrate momentum self-supervision and virtual categories into the Momentum Tightness and Contrast framework (MoTiC), constructing a feature space with rich representations and enhanced interclass cohesion. Experiments on three FSCIL benchmarks produce state-of-the-art performances, particularly on the fine-grained task CUB-200, validating our method's ability to reduce estimation bias and improve incremental learning robustness. |
| title | MoTiC: Momentum Tightness and Contrast for Few-Shot Class-Incremental Learning |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2509.19664 |