![\"\"](\"https://www.googleapis.com/download/storage/v1/b/kaggle-forum-message-attachments/o/inbox%2F4461928%2F7f4ba05c4be05e7a42d3c55412e8fda1%2Fgithub.jpg?generation=1690520546794678&alt=media\")
I've created a repository that might be helpful to those participating in the competition.
\nThis repository is designed to aid participants in testing their models in a more interactive and intuitive way, using live webcam input or any other videos. The interface is built using Gradio and it should be fairly simple to integrate your own model for testing.
\nTo use it, you just need to place your TensorFlow Lite (.tflite) model and an inference_args.json file in the weights directory. Then, after setting up the environment (a straightforward process, as described in the README file), you can run the Gradio app, which will accept webcam input/video upload. Just record (or upload) a short clip and press the submit button, and you will get the predictions in some time.
I've created a repository that might be helpful to those participating in the competition.
\nThis repository is designed to aid participants in testing their models in a more interactive and intuitive way, using live webcam input or any other videos. The interface is built using Gradio and it should be fairly simple to integrate your own model for testing.
\nTo use it, you just need to place your TensorFlow Lite (.tflite) model and an inference_args.json file in the weights directory. Then, after setting up the environment (a straightforward process, as described in the README file), you can run the Gradio app, which will accept webcam input/video upload. Just record (or upload) a short clip and press the submit button, and you will get the predictions in some time.
Great idea! I'm trying to do a similar approach on an Android device. We'll see how that goes. ;}
", "rawMarkdown": "Great idea! I'm trying to do a similar approach on an Android device. We'll see how that goes. ;}", "votes": null }, { "id": "2383952", "postDate": "08/10/2023 17:07:07", "content": "Hello! I have been trying to do this as well, but I am having little luck. Were you able to make any progress on it?
", "rawMarkdown": "Hello! I have been trying to do this as well, but I am having little luck. Were you able to make any progress on it?", "votes": null } ], "comments": [ { "id": 2369196, "author_name": "matucker", "author_url": "", "post_date": "08/01/2023 15:23:53", "content": "Great idea! I'm trying to do a similar approach on an Android device. We'll see how that goes. ;}
", "votes": null, "replies": [ { "id": 2383952, "author_name": "ethanscheys", "author_url": "", "post_date": "08/10/2023 17:07:07", "content": "Hello! I have been trying to do this as well, but I am having little luck. Were you able to make any progress on it?
", "votes": null, "replies": [] } ] } ], "raw_markdown_by_id": { "2362421": "I've created a repository that might be helpful to those participating in the competition.\n\n[This repository ](https://github.com/SamratThapa120/gradio-ASL-fingerspelling-recognition)is designed to aid participants in testing their models in a more interactive and intuitive way, using live webcam input or any other videos. The interface is built using Gradio and it should be fairly simple to integrate your own model for testing.\n\nTo use it, you just need to place your TensorFlow Lite (.tflite) model and an inference_args.json file in the `weights` directory. Then, after setting up the environment (a straightforward process, as described in the README file), you can run the Gradio app, which will accept webcam input/video upload. Just record (or upload) a short clip and press the submit button, and you will get the predictions in some time. \n\n ", "2369196": "Great idea! I'm trying to do a similar approach on an Android device. We'll see how that goes. ;}", "2383952": "Hello! I have been trying to do this as well, but I am having little luck. Were you able to make any progress on it?" }, "source": "meta" }