Conference paper Open Access

Incorporation of Semantic Segmentation Information in Deep Hashing Techniques for Image Retrieval

Gkountakos Konstantinos; Semertzidis Theodoros; Papadopoulos Georgios Th.; Daras Petros


JSON Export

{
  "files": [
    {
      "links": {
        "self": "https://zenodo.org/api/files/f3155632-9de5-414e-b10c-a196901c4dc6/IEEE-ICE2017-258.pdf"
      }, 
      "checksum": "md5:65094795f4291154cd3bdf6050ed3e3a", 
      "bucket": "f3155632-9de5-414e-b10c-a196901c4dc6", 
      "key": "IEEE-ICE2017-258.pdf", 
      "type": "pdf", 
      "size": 1721177
    }
  ], 
  "owners": [
    19291
  ], 
  "doi": "10.5281/zenodo.1076432", 
  "stats": {
    "version_unique_downloads": 37.0, 
    "unique_views": 96.0, 
    "views": 96.0, 
    "version_views": 96.0, 
    "unique_downloads": 37.0, 
    "version_unique_views": 96.0, 
    "volume": 65404726.0, 
    "version_downloads": 38.0, 
    "downloads": 38.0, 
    "version_volume": 65404726.0
  }, 
  "links": {
    "doi": "https://doi.org/10.5281/zenodo.1076432", 
    "conceptdoi": "https://doi.org/10.5281/zenodo.1076431", 
    "bucket": "https://zenodo.org/api/files/f3155632-9de5-414e-b10c-a196901c4dc6", 
    "conceptbadge": "https://zenodo.org/badge/doi/10.5281/zenodo.1076431.svg", 
    "html": "https://zenodo.org/record/1076432", 
    "latest_html": "https://zenodo.org/record/1076432", 
    "badge": "https://zenodo.org/badge/doi/10.5281/zenodo.1076432.svg", 
    "latest": "https://zenodo.org/api/records/1076432"
  }, 
  "conceptdoi": "10.5281/zenodo.1076431", 
  "created": "2017-12-04T14:25:07.315509+00:00", 
  "updated": "2020-01-20T15:11:55.779410+00:00", 
  "conceptrecid": "1076431", 
  "revision": 5, 
  "id": 1076432, 
  "metadata": {
    "access_right_category": "success", 
    "doi": "10.5281/zenodo.1076432", 
    "version": "Camera ready", 
    "language": "eng", 
    "title": "Incorporation of Semantic Segmentation Information in Deep Hashing Techniques for Image Retrieval", 
    "license": {
      "id": "CC-BY-4.0"
    }, 
    "related_identifiers": [
      {
        "scheme": "doi", 
        "identifier": "10.5281/zenodo.1076431", 
        "relation": "isVersionOf"
      }
    ], 
    "relations": {
      "version": [
        {
          "count": 1, 
          "index": 0, 
          "parent": {
            "pid_type": "recid", 
            "pid_value": "1076431"
          }, 
          "is_last": true, 
          "last_child": {
            "pid_type": "recid", 
            "pid_value": "1076432"
          }
        }
      ]
    }, 
    "grants": [
      {
        "code": "700367", 
        "links": {
          "self": "https://zenodo.org/api/grants/10.13039/501100000780::700367"
        }, 
        "title": "Detecting and ANalysing TErrorist-related online contents and financing activities", 
        "acronym": "DANTE", 
        "program": "H2020", 
        "funder": {
          "doi": "10.13039/501100000780", 
          "acronyms": [], 
          "name": "European Commission", 
          "links": {
            "self": "https://zenodo.org/api/funders/10.13039/501100000780"
          }
        }
      }
    ], 
    "keywords": [
      "Hashing", 
      "Deep learning", 
      "Image retrieval"
    ], 
    "publication_date": "2017-06-29", 
    "creators": [
      {
        "affiliation": "Centre for Research and Technology Hellas", 
        "name": "Gkountakos Konstantinos"
      }, 
      {
        "affiliation": "Centre for Research and Technology Hellas", 
        "name": "Semertzidis Theodoros"
      }, 
      {
        "affiliation": "Centre for Research and Technology Hellas", 
        "name": "Papadopoulos Georgios Th."
      }, 
      {
        "affiliation": "Centre for Research and Technology Hellas", 
        "name": "Daras Petros"
      }
    ], 
    "access_right": "open", 
    "resource_type": {
      "subtype": "conferencepaper", 
      "type": "publication", 
      "title": "Conference paper"
    }, 
    "description": "<p>Extracting discriminative image features for similarity search in nowadays large-scale databases becomes an imperative issue of paramount importance. To address the so called task of Approximate Nearest Neighbor (ANN) search in large visual dataset, deep hashing methods (i.e. approaches that make use of the recent deep learning paradigm in computer vision) have recently been introduced. In this paper, a novel approach to deep hashing is proposed, which incorporates local-level information, in the form of image semantic segmentation masks, during the hash code learning step. The proposed framework makes use of pixel-level classification labels, i.e. following a point-wise supervised learning methodology. Experimental evaluation in the significantly challenging domain of on-line terrorist propaganda video analysis, i.e. a highly diverse and heterogeneous application case, demonstrates the efficiency of the proposed approach.</p>"
  }
}
96
38
views
downloads
All versions This version
Views 9696
Downloads 3838
Data volume 65.4 MB65.4 MB
Unique views 9696
Unique downloads 3737

Share

Cite as