[2604.02765] Towards Realistic Class-Incremental Learning with Free-Flow Increments

Computer Science > Machine Learning arXiv:2604.02765 (cs) [Submitted on 3 Apr 2026] Title:Towards Realistic Class-Incremental Learning with Free-Flow Increments Authors:Zhiming Xu, Baile Xu, Jian Zhao, Furao Shen, Suorong Yang View a PDF of the paper titled Towards Realistic Class-Incremental Learning with Free-Flow Increments, by Zhiming Xu and 4 other authors View PDF HTML (experimental) Abstract:Class-incremental learning (CIL) is typically evaluated under predefined schedules with equal-sized tasks, leaving more realistic and complex cases unexplored. However, a practical CIL system should learns immediately when any number of new classes arrive, without forcing fixed-size tasks. We formalize this setting as Free-Flow Class-Incremental Learning (FFCIL), where data arrives as a more realistic stream with a highly variable number of unseen classes each step. It will make many existing CIL methods brittle and lead to clear performance degradation. We propose a model-agnostic framework for robust CIL learning under free-flow arrivals. It comprises a class-wise mean (CWM) objective that replaces sample frequency weighted loss with uniformly aggregated class-conditional supervision, thereby stabilizing the learning signal across free-flow class increments, as well as method-wise adjustments that improve robustness for representative CIL paradigms. Specifically, we constrain distillation to replayed data, normalize the scale of contrastive and knowledge transfer losses, and i...