{
  "id": "2311.13718",
  "version": "v1",
  "published": "2023-11-22T22:23:34.000Z",
  "updated": "2023-11-22T22:23:34.000Z",
  "title": "A Unified Approach to Count-Based Weakly-Supervised Learning",
  "authors": [
    "Vinay Shukla",
    "Zhe Zeng",
    "Kareem Ahmed",
    "Guy Van den Broeck"
  ],
  "categories": [
    "cs.LG",
    "cs.AI"
  ],
  "abstract": "High-quality labels are often very scarce, whereas unlabeled data with inferred weak labels occurs more naturally. In many cases, these weak labels dictate the frequency of each respective class over a set of instances. In this paper, we develop a unified approach to learning from such weakly-labeled data, which we call count-based weakly-supervised learning. At the heart of our approach is the ability to compute the probability of exactly k out of n outputs being set to true. This computation is differentiable, exact, and efficient. Building upon the previous computation, we derive a count loss penalizing the model for deviations in its distribution from an arithmetic constraint defined over label counts. We evaluate our approach on three common weakly-supervised learning paradigms and observe that our proposed approach achieves state-of-the-art or highly competitive results across all three of the paradigms.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2023-11-22T22:23:34.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "count-based weakly-supervised learning",
      "unified approach",
      "approach achieves state-of-the-art",
      "inferred weak labels occurs",
      "weak labels dictate"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}