SAMix: Calibrated and Accurate Continual Learning via Sphere-Adaptive Mixup and Neural Collapse
This work addresses the need for more reliable and calibrated predictions in continual learning systems, representing an incremental improvement over existing neural collapse-based methods.
The paper tackles the problem of improving both accuracy and calibration in continual learning by proposing Sphere-Adaptive Mixup (SAMix), which adapts mixup to neural collapse geometry, resulting in significant performance boosts and enhanced model reliability.
While most continual learning methods focus on mitigating forgetting and improving accuracy, they often overlook the critical aspect of network calibration, despite its importance. Neural collapse, a phenomenon where last-layer features collapse to their class means, has demonstrated advantages in continual learning by reducing feature-classifier misalignment. Few works aim to improve the calibration of continual models for more reliable predictions. Our work goes a step further by proposing a novel method that not only enhances calibration but also improves performance by reducing overconfidence, mitigating forgetting, and increasing accuracy. We introduce Sphere-Adaptive Mixup (SAMix), an adaptive mixup strategy tailored for neural collapse-based methods. SAMix adapts the mixing process to the geometric properties of feature spaces under neural collapse, ensuring more robust regularization and alignment. Experiments show that SAMix significantly boosts performance, surpassing SOTA methods in continual learning while also improving model calibration. SAMix enhances both across-task accuracy and the broader reliability of predictions, making it a promising advancement for robust continual learning systems.