Queried Unlabeled Data Improves and Robustifies Class-Incremental Learning. (arXiv:2206.07842v2 [cs.LG] UPDATED)

Web: http://arxiv.org/abs/2206.07842

Class-incremental learning (CIL) suffers from the notorious dilemma between
learning newly added classes and preserving previously learned class knowledge.
That catastrophic forgetting issue could be mitigated by storing historical
data for replay, which yet would cause memory overheads as well as imbalanced
prediction updates. To address this dilemma, we propose to leverage "free"
external unlabeled data querying in continual learning. We first present a CIL
with Queried Unlabeled Data (CIL-QUD) scheme, where we only store a handful of
past training …

arxiv data incremental learning lg