[2506.19558] Consistency-Driven Calibration and Matching for Few-Shot Class-Incremental Learning

Computer Science > Machine Learning arXiv:2506.19558 (cs) [Submitted on 24 Jun 2025 (v1), last revised 1 Mar 2026 (this version, v2)] Title:Consistency-Driven Calibration and Matching for Few-Shot Class-Incremental Learning Authors:Qinzhe Wang, Zixuan Chen, Keke Huang, Xiu Su, Chunhua Yang, Chang Xu View a PDF of the paper titled Consistency-Driven Calibration and Matching for Few-Shot Class-Incremental Learning, by Qinzhe Wang and 5 other authors View PDF HTML (experimental) Abstract:Few-Shot Class Incremental Learning (FSCIL) is crucial for adapting to the complex open-world environments. Contemporary prospective learning-based space construction methods struggle to balance old and new knowledge, as prototype bias and rigid structures limit the expressive capacity of the embedding space. Different from these strategies, we rethink the optimization dilemma from the perspective of feature-structure dual consistency, and propose a Consistency-driven Calibration and Matching (ConCM) framework that systematically mitigates the knowledge conflict inherent in FSCIL. Specifically, inspired by hippocampal associative memory, we design a memory-aware prototype calibration that extracts generalized semantic attributes from base classes and reintegrates them into novel classes to enhance the conceptual center consistency of features. Further, to consolidate memory associations, we propose dynamic structure matching, which adaptively aligns the calibrated features to a session-specif...