Conference paper Open Access

Near-Duplicate Video Retrieval with Deep Metric Learning

Kordopatis-Zilos, Giorgos; Papadopoulos, Symeon; Patras, Ioannis; Kompatsiaris, Yiannis


JSON Export

{
  "files": [
    {
      "links": {
        "self": "https://zenodo.org/api/files/dbf6682e-e1ff-4a0a-a9cc-eee8961cd0a9/duplicate_video_retrieval_camera_ready.pdf"
      }, 
      "checksum": "md5:a50b70d87c96310db1b060b4dbf93329", 
      "bucket": "dbf6682e-e1ff-4a0a-a9cc-eee8961cd0a9", 
      "key": "duplicate_video_retrieval_camera_ready.pdf", 
      "type": "pdf", 
      "size": 1384334
    }
  ], 
  "owners": [
    27159
  ], 
  "doi": "10.5281/zenodo.852351", 
  "stats": {
    "version_unique_downloads": 80.0, 
    "unique_views": 92.0, 
    "views": 98.0, 
    "downloads": 85.0, 
    "unique_downloads": 80.0, 
    "version_unique_views": 92.0, 
    "volume": 117668390.0, 
    "version_downloads": 85.0, 
    "version_views": 98.0, 
    "version_volume": 117668390.0
  }, 
  "links": {
    "doi": "https://doi.org/10.5281/zenodo.852351", 
    "conceptdoi": "https://doi.org/10.5281/zenodo.852350", 
    "bucket": "https://zenodo.org/api/files/dbf6682e-e1ff-4a0a-a9cc-eee8961cd0a9", 
    "conceptbadge": "https://zenodo.org/badge/doi/10.5281/zenodo.852350.svg", 
    "html": "https://zenodo.org/record/852351", 
    "latest_html": "https://zenodo.org/record/852351", 
    "badge": "https://zenodo.org/badge/doi/10.5281/zenodo.852351.svg", 
    "latest": "https://zenodo.org/api/records/852351"
  }, 
  "conceptdoi": "10.5281/zenodo.852350", 
  "created": "2017-08-28T11:09:46.161509+00:00", 
  "updated": "2019-04-10T04:19:00.414878+00:00", 
  "conceptrecid": "852350", 
  "revision": 4, 
  "id": 852351, 
  "metadata": {
    "access_right_category": "success", 
    "doi": "10.5281/zenodo.852351", 
    "description": "<p>This work addresses the problem of Near-Duplicate Video Retrieval (NDVR). We propose an effective video-level NDVR scheme based on deep metric learning that leverages Convolutional Neural Network (CNN) features from intermediate layers to generate discriminative global video representations in tandem with a Deep Metric Learning (DML) framework with two fusion variations, trained to approximate an embedding function for accurate distance calculation between two near-duplicate videos. In contrast to most state-of-the-art methods, which exploit information deriving from the same source of data for both development and evaluation (which usually results to dataset-specific solutions), the proposed model is fed during training with sampled triplets generated from an independent dataset and is thoroughly tested on the widely used CC WEB VIDEO dataset, using two popular deep CNN architectures (AlexNet, GoogleNet). We demonstrate that the proposed approach achieves outstanding performance against the state-of-the-art, either with or without access to the evaluation dataset.</p>", 
    "license": {
      "id": "CC-BY-4.0"
    }, 
    "title": "Near-Duplicate Video Retrieval with Deep Metric Learning", 
    "relations": {
      "version": [
        {
          "count": 1, 
          "index": 0, 
          "parent": {
            "pid_type": "recid", 
            "pid_value": "852350"
          }, 
          "is_last": true, 
          "last_child": {
            "pid_type": "recid", 
            "pid_value": "852351"
          }
        }
      ]
    }, 
    "communities": [
      {
        "id": "invid-h2020"
      }
    ], 
    "grants": [
      {
        "code": "687786", 
        "links": {
          "self": "https://zenodo.org/api/grants/10.13039/501100000780::687786"
        }, 
        "title": "In Video Veritas \u2013 Verification of Social Media Video Content for the News Industry", 
        "acronym": "InVID", 
        "program": "H2020", 
        "funder": {
          "doi": "10.13039/501100000780", 
          "acronyms": [
            "EC"
          ], 
          "name": "European Commission", 
          "links": {
            "self": "https://zenodo.org/api/funders/10.13039/501100000780"
          }
        }
      }
    ], 
    "keywords": [
      "Near-Duplicate", 
      "Video Retrieval", 
      "Deep Metric Learning"
    ], 
    "publication_date": "2017-10-23", 
    "creators": [
      {
        "affiliation": "CERTH-ITI", 
        "name": "Kordopatis-Zilos, Giorgos"
      }, 
      {
        "affiliation": "CERTH-ITI", 
        "name": "Papadopoulos, Symeon"
      }, 
      {
        "affiliation": "CERTH-ITI", 
        "name": "Patras, Ioannis"
      }, 
      {
        "affiliation": "CERTH-ITI", 
        "name": "Kompatsiaris, Yiannis"
      }
    ], 
    "meeting": {
      "acronym": "ICCV 2017", 
      "url": "http://iccv2017.thecvf.com/", 
      "dates": "October 23 2017", 
      "place": "Venice, Italy", 
      "title": "5th Workshop on Web-scale Vision and Social Media"
    }, 
    "access_right": "open", 
    "resource_type": {
      "subtype": "conferencepaper", 
      "type": "publication", 
      "title": "Conference paper"
    }, 
    "related_identifiers": [
      {
        "scheme": "doi", 
        "identifier": "10.5281/zenodo.852350", 
        "relation": "isVersionOf"
      }
    ]
  }
}
98
85
views
downloads
All versions This version
Views 9898
Downloads 8585
Data volume 117.7 MB117.7 MB
Unique views 9292
Unique downloads 8080

Share

Cite as