{
  "id": 201963,
  "title": "Have anyone measured ADE \\ FDE metrics?",
  "url": "/competitions/lyft-motion-prediction-autonomous-vehicles/discussion/201963",
  "author_name": "Aleksey",
  "post_date": "2020-12-07T14:43:05.824000",
  "votes": 1,
  "comment_count": 0,
  "views": 0,
  "content": "<p>I am wondering have anyone measured Average Displacement Error and Final Displacement Error of theirs models?</p>\n<p>It seems that in many papers the most popular metrics are ADE\\FDE\\NLL.<br>\nIn case of this competition provided metric is related to NLL, but it is interesting how accurate model in terms of euclidean errors.</p>\n<p>[pic from Rudenko, Andrey, et al. \"Human motion trajectory prediction: A survey.\" The International Journal of Robotics Research 39.8 (2020): 895-935.]</p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F745756%2F1fa6c7b1f822c36c5dc625635a03cbeb%2FScreenshot%20from%202020-12-07%2017-40-16.png?generation=1607352263614157&amp;alt=media\" alt=\"\"></p>",
  "messages": [
    {
      "id": 1105131,
      "postDate": "2020-12-07T14:43:05.823Z",
      "content": "<p>I am wondering have anyone measured Average Displacement Error and Final Displacement Error of theirs models?</p>\n<p>It seems that in many papers the most popular metrics are ADE\\FDE\\NLL.<br>\nIn case of this competition provided metric is related to NLL, but it is interesting how accurate model in terms of euclidean errors.</p>\n<p>[pic from Rudenko, Andrey, et al. \"Human motion trajectory prediction: A survey.\" The International Journal of Robotics Research 39.8 (2020): 895-935.]</p>\n<p><img src=\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F745756%2F1fa6c7b1f822c36c5dc625635a03cbeb%2FScreenshot%20from%202020-12-07%2017-40-16.png?generation=1607352263614157&amp;alt=media\" alt=\"\"></p>",
      "rawMarkdown": "I am wondering have anyone measured Average Displacement Error and Final Displacement Error of theirs models?\n\nIt seems that in many papers the most popular metrics are ADE\\FDE\\NLL.\nIn case of this competition provided metric is related to NLL, but it is interesting how accurate model in terms of euclidean errors.\n\n[pic from Rudenko, Andrey, et al. \"Human motion trajectory prediction: A survey.\" The International Journal of Robotics Research 39.8 (2020): 895-935.]\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F745756%2F1fa6c7b1f822c36c5dc625635a03cbeb%2FScreenshot%20from%202020-12-07%2017-40-16.png?generation=1607352263614157&alt=media)",
      "votes": 1
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "1105131": "I am wondering have anyone measured Average Displacement Error and Final Displacement Error of theirs models?\n\nIt seems that in many papers the most popular metrics are ADE\\FDE\\NLL.\nIn case of this competition provided metric is related to NLL, but it is interesting how accurate model in terms of euclidean errors.\n\n[pic from Rudenko, Andrey, et al. \"Human motion trajectory prediction: A survey.\" The International Journal of Robotics Research 39.8 (2020): 895-935.]\n\n![](https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F745756%2F1fa6c7b1f822c36c5dc625635a03cbeb%2FScreenshot%20from%202020-12-07%2017-40-16.png?generation=1607352263614157&alt=media)"
  }
}