{
  "id": 54730,
  "title": "Instance Segmentation is Awesome!",
  "url": "/competitions/data-science-bowl-2018/discussion/54730",
  "author_name": "JamesEverard",
  "post_date": "2018-04-17T06:49:47.902000",
  "votes": 8,
  "comment_count": 0,
  "views": 0,
  "content": "<p>I was just reflecting on how far we have come since the sea lion counting competition this time last year.</p>\n\n<p>Probably the top 100 people in this nucleus competition would have beaten the top score in the sea lion competition using the code they have written over the last three months (I could be wrong). </p>\n\n<p>And I look at the results that my borrowed ordinary mask-RCNN code produces and it just go 'WOW! I can't believe I have written code that does THIS!'</p>\n\n<p>For me deep learning in perceptual systems like Vision and Speech recognition is easily our most impressive achievement in AI. I mean, two years ago, could you have imagined that you would actually be writing code this clever?</p>\n\n<p>AI Chat bots are just mimicry. NLP just manipulating symbols.</p>\n\n<p>But my eyes and my ears are looking like seriously out-of-date technology.</p>\n\n<p>Instance Segmentation is awesome!</p>",
  "messages": [
    {
      "id": 315480,
      "postDate": "2018-04-17T06:49:47.903Z",
      "content": "<p>I was just reflecting on how far we have come since the sea lion counting competition this time last year.</p>\n\n<p>Probably the top 100 people in this nucleus competition would have beaten the top score in the sea lion competition using the code they have written over the last three months (I could be wrong). </p>\n\n<p>And I look at the results that my borrowed ordinary mask-RCNN code produces and it just go 'WOW! I can't believe I have written code that does THIS!'</p>\n\n<p>For me deep learning in perceptual systems like Vision and Speech recognition is easily our most impressive achievement in AI. I mean, two years ago, could you have imagined that you would actually be writing code this clever?</p>\n\n<p>AI Chat bots are just mimicry. NLP just manipulating symbols.</p>\n\n<p>But my eyes and my ears are looking like seriously out-of-date technology.</p>\n\n<p>Instance Segmentation is awesome!</p>",
      "rawMarkdown": "I was just reflecting on how far we have come since the sea lion counting competition this time last year.\n\nProbably the top 100 people in this nucleus competition would have beaten the top score in the sea lion competition using the code they have written over the last three months (I could be wrong). \n\nAnd I look at the results that my borrowed ordinary mask-RCNN code produces and it just go 'WOW! I can't believe I have written code that does THIS!'\n\nFor me deep learning in perceptual systems like Vision and Speech recognition is easily our most impressive achievement in AI. I mean, two years ago, could you have imagined that you would actually be writing code this clever?\n\nAI Chat bots are just mimicry. NLP just manipulating symbols.\n\nBut my eyes and my ears are looking like seriously out-of-date technology.\n\nInstance Segmentation is awesome!",
      "votes": 8
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "315480": "I was just reflecting on how far we have come since the sea lion counting competition this time last year.\n\nProbably the top 100 people in this nucleus competition would have beaten the top score in the sea lion competition using the code they have written over the last three months (I could be wrong). \n\nAnd I look at the results that my borrowed ordinary mask-RCNN code produces and it just go 'WOW! I can't believe I have written code that does THIS!'\n\nFor me deep learning in perceptual systems like Vision and Speech recognition is easily our most impressive achievement in AI. I mean, two years ago, could you have imagined that you would actually be writing code this clever?\n\nAI Chat bots are just mimicry. NLP just manipulating symbols.\n\nBut my eyes and my ears are looking like seriously out-of-date technology.\n\nInstance Segmentation is awesome!"
  }
}