Sculpting Margin Penalty: Intra-Task Adapter Merging and Classifier Calibration for Few-Shot Class-Incremental Learning

2508.05094v1 cs.CV 2025-08-09
Авторы:

Liang Bai, Hong Song, Jinfu Li, Yucong Lin, Jingfan Fan, Tianyu Fu, Danni Ai, Deqiang Xiao, Jian Yang

Резюме на русском

Нехватка данных для обучения в реальных приложениях часто приводит к значительному снижению качества методов классификации в задачах классового инкрементального обучения. Это особенно актуально для Few-Shot Class-Incremental Learning (FSCIL), где необходимо совмещать дискриминативность базовых классов и универсальность классификатора для новых классов. Большинство существующих подходов сталкиваются с проблемами в достижении этой балансировки. Мы предлагаем SMP (Sculpting Margin Penalty), новую методику FSCIL, которая использует параметрическую оптимизацию с механизмом Margin-aware Intra-task Adapter Merging (MIAM) для улучшения дискриминативности базовых классов и MPCC для калибровки классификатора на все видимые классы. Эксперименты на CIFAR-100, ImageNet-R и CUB-200 показали, что SMP достигает лидирующих результатов в FSCIL, сохраняя баланс между классами.

Abstract

Real-world applications often face data privacy constraints and high acquisition costs, making the assumption of sufficient training data in incremental tasks unrealistic and leading to significant performance degradation in class-incremental learning. Forward-compatible learning, which prospectively prepares for future tasks during base task training, has emerged as a promising solution for Few-Shot Class-Incremental Learning (FSCIL). However, existing methods still struggle to balance base-class discriminability and new-class generalization. Moreover, limited access to original data during incremental tasks often results in ambiguous inter-class decision boundaries. To address these challenges, we propose SMP (Sculpting Margin Penalty), a novel FSCIL method that strategically integrates margin penalties at different stages within the parameter-efficient fine-tuning paradigm. Specifically, we introduce the Margin-aware Intra-task Adapter Merging (MIAM) mechanism for base task learning. MIAM trains two sets of low-rank adapters with distinct classification losses: one with a margin penalty to enhance base-class discriminability, and the other without margin constraints to promote generalization to future new classes. These adapters are then adaptively merged to improve forward compatibility. For incremental tasks, we propose a Margin Penalty-based Classifier Calibration (MPCC) strategy to refine decision boundaries by fine-tuning classifiers on all seen classes' embeddings with a margin penalty. Extensive experiments on CIFAR100, ImageNet-R, and CUB200 demonstrate that SMP achieves state-of-the-art performance in FSCIL while maintaining a better balance between base and new classes.

Ссылки и действия