LGCVMar 26, 2024

DS-AL: A Dual-Stream Analytic Learning for Exemplar-Free Class-Incremental Learning

arXiv:2403.17503v153 citationsh-index: 8Has CodeAAAI
Originality Highly original
AI Analysis

This addresses the problem of catastrophic forgetting in class-incremental learning without storing past samples, offering a novel solution for scenarios where data privacy or storage constraints prevent replay.

The paper tackles exemplar-free class-incremental learning, which suffers from catastrophic forgetting, by proposing a dual-stream analytic learning approach that achieves performance comparable to or better than replay-based methods on datasets like CIFAR-100 and ImageNet, and demonstrates phase-invariance with identical performance in 500-phase and 5-phase tasks.

Class-incremental learning (CIL) under an exemplar-free constraint has presented a significant challenge. Existing methods adhering to this constraint are prone to catastrophic forgetting, far more so than replay-based techniques that retain access to past samples. In this paper, to solve the exemplar-free CIL problem, we propose a Dual-Stream Analytic Learning (DS-AL) approach. The DS-AL contains a main stream offering an analytical (i.e., closed-form) linear solution, and a compensation stream improving the inherent under-fitting limitation due to adopting linear mapping. The main stream redefines the CIL problem into a Concatenated Recursive Least Squares (C-RLS) task, allowing an equivalence between the CIL and its joint-learning counterpart. The compensation stream is governed by a Dual-Activation Compensation (DAC) module. This module re-activates the embedding with a different activation function from the main stream one, and seeks fitting compensation by projecting the embedding to the null space of the main stream's linear mapping. Empirical results demonstrate that the DS-AL, despite being an exemplar-free technique, delivers performance comparable with or better than that of replay-based methods across various datasets, including CIFAR-100, ImageNet-100 and ImageNet-Full. Additionally, the C-RLS' equivalent property allows the DS-AL to execute CIL in a phase-invariant manner. This is evidenced by a never-before-seen 500-phase CIL ImageNet task, which performs on a level identical to a 5-phase one. Our codes are available at https://github.com/ZHUANGHP/Analytic-continual-learning.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes