English

Continual Learning on Noisy Data Streams via Self-Purified Replay

Machine Learning 2021-10-18 v1

Abstract

Continually learning in the real world must overcome many challenges, among which noisy labels are a common and inevitable issue. In this work, we present a repla-ybased continual learning framework that simultaneously addresses both catastrophic forgetting and noisy labels for the first time. Our solution is based on two observations; (i) forgetting can be mitigated even with noisy labels via self-supervised learning, and (ii) the purity of the replay buffer is crucial. Building on this regard, we propose two key components of our method: (i) a self-supervised replay technique named Self-Replay which can circumvent erroneous training signals arising from noisy labeled data, and (ii) the Self-Centered filter that maintains a purified replay buffer via centrality-based stochastic graph ensembles. The empirical results on MNIST, CIFAR-10, CIFAR-100, and WebVision with real-world noise demonstrate that our framework can maintain a highly pure replay buffer amidst noisy streamed data while greatly outperforming the combinations of the state-of-the-art continual learning and noisy label learning methods. The source code is available at http://vision.snu.ac.kr/projects/SPR

Keywords

Cite

@article{arxiv.2110.07735,
  title  = {Continual Learning on Noisy Data Streams via Self-Purified Replay},
  author = {Chris Dongjoo Kim and Jinseo Jeong and Sangwoo Moon and Gunhee Kim},
  journal= {arXiv preprint arXiv:2110.07735},
  year   = {2021}
}

Comments

Published at ICCV 2021 main conference

R2 v1 2026-06-24T06:54:13.712Z