{
  "id": 221070,
  "title": "UP-DETR: Unsupervised Pre-training for Object Detection with Transformers",
  "url": "/competitions/rfcx-species-audio-detection/discussion/221070",
  "author_name": "hengck23",
  "post_date": "2021-02-21T00:45:34.733000",
  "votes": 11,
  "comment_count": 0,
  "views": 0,
  "content": "<p>i just read paper and find actually it can be relevant to this challenge</p>\n<p>QUOTE:<br>\nThis procedure can be treated as unsupervised one-shot detection or deep learning based template matching.</p>\n<p><img src=\"https://pbs.twimg.com/media/EnJv4n3WMAIfL8P.png\" alt=\"\"></p>\n<p>in theory:</p>\n<ul>\n<li>query = annotated tp</li>\n<li>prediction best-matched box. Note that we should modify from the paper so that query (reference template plate) don't have to come from the same input image</li>\n</ul>\n<p>related:<br>\nFew-Shot Object Detection with Attention-RPN and Multi-Relation Detector<br>\n<a href=\"https://arxiv.org/pdf/1908.01998v1.pdf\" target=\"_blank\">https://arxiv.org/pdf/1908.01998v1.pdf</a></p>\n<p><img src=\"https://d3i71xaburhd42.cloudfront.net/58d20519b190308d144d4b0110527231179ea100/1-Figure1-1.png\" alt=\"\"></p>\n<p>2 boxes at the top of image is the input query template</p>",
  "messages": [
    {
      "id": 1212173,
      "postDate": "2021-02-21T00:45:34.733Z",
      "content": "<p>i just read paper and find actually it can be relevant to this challenge</p>\n<p>QUOTE:<br>\nThis procedure can be treated as unsupervised one-shot detection or deep learning based template matching.</p>\n<p><img src=\"https://pbs.twimg.com/media/EnJv4n3WMAIfL8P.png\" alt=\"\"></p>\n<p>in theory:</p>\n<ul>\n<li>query = annotated tp</li>\n<li>prediction best-matched box. Note that we should modify from the paper so that query (reference template plate) don't have to come from the same input image</li>\n</ul>\n<p>related:<br>\nFew-Shot Object Detection with Attention-RPN and Multi-Relation Detector<br>\n<a href=\"https://arxiv.org/pdf/1908.01998v1.pdf\" target=\"_blank\">https://arxiv.org/pdf/1908.01998v1.pdf</a></p>\n<p><img src=\"https://d3i71xaburhd42.cloudfront.net/58d20519b190308d144d4b0110527231179ea100/1-Figure1-1.png\" alt=\"\"></p>\n<p>2 boxes at the top of image is the input query template</p>",
      "rawMarkdown": "i just read paper and find actually it can be relevant to this challenge\n\nQUOTE:\nThis procedure can be treated as unsupervised one-shot detection or deep learning based template matching.\n\n![](https://pbs.twimg.com/media/EnJv4n3WMAIfL8P.png)\n\nin theory:\n- query = annotated tp\n- prediction best-matched box. Note that we should modify from the paper so that query (reference template plate) don't have to come from the same input image\n\n\nrelated:\nFew-Shot Object Detection with Attention-RPN and Multi-Relation Detector\nhttps://arxiv.org/pdf/1908.01998v1.pdf\n\n![](https://d3i71xaburhd42.cloudfront.net/58d20519b190308d144d4b0110527231179ea100/1-Figure1-1.png)\n\n2 boxes at the top of image is the input query template",
      "votes": 10
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1212173": "i just read paper and find actually it can be relevant to this challenge\n\nQUOTE:\nThis procedure can be treated as unsupervised one-shot detection or deep learning based template matching.\n\n![](https://pbs.twimg.com/media/EnJv4n3WMAIfL8P.png)\n\nin theory:\n- query = annotated tp\n- prediction best-matched box. Note that we should modify from the paper so that query (reference template plate) don't have to come from the same input image\n\n\nrelated:\nFew-Shot Object Detection with Attention-RPN and Multi-Relation Detector\nhttps://arxiv.org/pdf/1908.01998v1.pdf\n\n![](https://d3i71xaburhd42.cloudfront.net/58d20519b190308d144d4b0110527231179ea100/1-Figure1-1.png)\n\n2 boxes at the top of image is the input query template"
  }
}