Integrating Present and Past in Unsupervised Continual Learning

arXiv:2404.19132v1 Announce Type: new
Abstract: We formulate a unifying framework for unsupervised continual learning (UCL), which disentangles learning objectives that are specific to the present and the past data, encompassing stability, plasticity, and cross-task consolidation. The framework reveals that many existing UCL approaches overlook cross-task consolidation and try to balance plasticity and stability in a shared embedding space. This results in worse performance due to a lack of within-task data diversity and reduced effectiveness in learning the current task. Our …

abstract arxiv balance consolidation continual cs.cv cs.lg data framework stability type unsupervised