GUIDE: Guidance-based Incremental Learning with Diffusion Models

arXiv:2403.03938v1 Announce Type: new
Abstract: We introduce GUIDE, a novel continual learning approach that directs diffusion models to rehearse samples at risk of being forgotten. Existing generative strategies combat catastrophic forgetting by randomly sampling rehearsal examples from a generative model. Such an approach contradicts buffer-based approaches where sampling strategy plays an important role. We propose to bridge this gap by integrating diffusion models with classifier guidance techniques to produce rehearsal examples specifically targeting information forgotten by a continuously trained model. …

abstract arxiv catastrophic forgetting continual cs.lg diffusion diffusion models examples generative guidance guide incremental novel risk role samples sampling strategies strategy type