Advancing Analytic Class-Incremental Learning through Vision-Language Calibration

TL;DR

This paper introduces VILA, a dual-branch framework that enhances class-incremental learning with pre-trained models by combining vision-language calibration to improve stability without sacrificing efficiency.

Contribution

VILA advances analytic class-incremental learning by integrating geometric and semantic calibration strategies, addressing representation rigidity and prediction bias.

Findings

VILA outperforms existing methods across eight benchmarks.

It maintains efficiency while improving stability in fine-grained scenarios.

The framework effectively combines feature-level and decision-level calibration.

Abstract

Class-incremental learning (CIL) with pre-trained models (PTMs) faces a critical trade-off between efficient adaptation and long-term stability. While analytic learning enables rapid, recursive closed-form updates, its efficacy is often compromised by accumulated errors and feature incompatibility. In this paper, we first conduct a systematic study to dissect the failure modes of PTM-based analytic CIL, identifying representation rigidity as the primary bottleneck. Motivated by this insight, we propose VILA, a novel dual-branch framework that advances analytic CIL via a two-level vision-language calibration strategy. Specifically, we coherently fuse plastic, task-adapted features with a frozen, universal visual anchor at the feature level through geometric calibration, and leverage cross-modal semantic priors at the decision level to rectify prediction bias. This confluence maintains analytic-learning's extreme efficiency while overcoming its inherent brittleness. Extensive experiments across eight benchmarks demonstrate that VILA consistently yields superior performance, particularly in fine-grained and long-sequence scenarios. Our framework harmonizes high-fidelity prediction with the simplicity of analytic learning. Our code is available at https://github.com/byzhaoAI/VILA.