{
  "id": "2004.00909",
  "version": "v1",
  "published": "2020-04-02T09:56:03.000Z",
  "updated": "2020-04-02T09:56:03.000Z",
  "title": "Learning Representations For Images With Hierarchical Labels",
  "authors": [
    "Ankit Dhall"
  ],
  "comment": "Master thesis",
  "categories": [
    "cs.LG",
    "cs.CV",
    "stat.ML"
  ],
  "abstract": "Image classification has been studied extensively but there has been limited work in the direction of using non-conventional, external guidance other than traditional image-label pairs to train such models. In this thesis we present a set of methods to leverage information about the semantic hierarchy induced by class labels. In the first part of the thesis, we inject label-hierarchy knowledge to an arbitrary classifier and empirically show that availability of such external semantic information in conjunction with the visual semantics from images boosts overall performance. Taking a step further in this direction, we model more explicitly the label-label and label-image interactions by using order-preserving embedding-based models, prevalent in natural language, and tailor them to the domain of computer vision to perform image classification. Although, contrasting in nature, both the CNN-classifiers injected with hierarchical information, and the embedding-based models outperform a hierarchy-agnostic model on the newly presented, real-world ETH Entomological Collection image dataset.",
  "revisions": [
    {
      "version": "v1",
      "updated": "2020-04-02T09:56:03.000Z"
    }
  ],
  "analyses": {
    "keywords": [
      "learning representations",
      "hierarchical labels",
      "eth entomological collection image dataset",
      "real-world eth entomological collection image",
      "images boosts overall performance"
    ],
    "tags": [
      "dissertation"
    ],
    "note": {
      "typesetting": "TeX",
      "pages": 0,
      "language": "en",
      "license": "arXiv",
      "status": "editable"
    }
  }
}