{
  "id": "1909.09338",
  "version": "v1",
  "published": "2019-09-20T06:15:13.000Z",
  "updated": "2019-09-20T06:15:13.000Z",
  "title": "A Simple yet Effective Baseline for Robust Deep Learning with Noisy Labels",
  "authors": [
    "Yucen Luo",
    "Jun Zhu",
    "Tomas Pfister"
  ],
  "categories": [
    "cs.LG",
    "stat.ML"
  ],
  "abstract": "Recently deep neural networks have shown their capacity to memorize training data, even with noisy labels, which hurts generalization performance. To mitigate this issue, we propose a simple but effective baseline that is robust to noisy labels, even with severe noise. Our objective involves a variance regularization term that implicitly penalizes the Jacobian norm of the neural network on the whole training set (including the noisy-labeled data), which encourages generalization and prevents overfitting to the corrupted labels. Experiments on both synthetically generated incorrect labels and realistic large-scale noisy datasets demonstrate that our approach achieves state-of-the-art performance with a high tolerance to severe noise.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2019-09-20T06:15:13.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "noisy labels",
      "robust deep learning",
      "effective baseline",
      "realistic large-scale noisy datasets demonstrate",
      "neural network"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}