{
  "id": 402505,
  "title": "Can we extend the tflite model to predict full sentences?",
  "url": "/competitions/asl-signs/discussion/402505",
  "author_name": "Dhanesh123",
  "post_date": "2023-04-18T16:53:08.847000",
  "votes": 1,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Hello Kagglers,<br>\nThis competition aims to detect single signs, but I was wondering if we could use this to predict a whole sentence using this same model without training on new data. I made a script to get the mediapipe coordinates from video and run the tflite model on it and get the predictions. I signed a whole word and tried to see if the signs used are in top 5 predictions but that didn't work. Can someone give an idea on how to achieve this?</p>",
  "messages": [
    {
      "id": 2226110,
      "postDate": "2023-04-18T16:53:08.847Z",
      "content": "<p>Hello Kagglers,<br>\nThis competition aims to detect single signs, but I was wondering if we could use this to predict a whole sentence using this same model without training on new data. I made a script to get the mediapipe coordinates from video and run the tflite model on it and get the predictions. I signed a whole word and tried to see if the signs used are in top 5 predictions but that didn't work. Can someone give an idea on how to achieve this?</p>",
      "rawMarkdown": "Hello Kagglers,\nThis competition aims to detect single signs, but I was wondering if we could use this to predict a whole sentence using this same model without training on new data. I made a script to get the mediapipe coordinates from video and run the tflite model on it and get the predictions. I signed a whole word and tried to see if the signs used are in top 5 predictions but that didn't work. Can someone give an idea on how to achieve this?",
      "votes": 1
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "2226110": "Hello Kagglers,\nThis competition aims to detect single signs, but I was wondering if we could use this to predict a whole sentence using this same model without training on new data. I made a script to get the mediapipe coordinates from video and run the tflite model on it and get the predictions. I signed a whole word and tried to see if the signs used are in top 5 predictions but that didn't work. Can someone give an idea on how to achieve this?"
  }
}