{
  "id": 498273,
  "title": "Bengali.AI-IUT Regional Speech Recognition Competition: Bashabichitra Results",
  "url": "/competitions/ben10/discussion/498273",
  "author_name": "Sushmit",
  "post_date": "2024-04-27T15:24:56.680000",
  "votes": 3,
  "comment_count": 0,
  "views": 0,
  "content": "<p>Greetings from Bengali.AI</p>\n<p>We thank the research community for participating in the competition and present their approaches to tackle Bengali Regional Speech Recognition. We congratulate the top 3 teams:  Jibonanondo Das, কাকাতুয়া(kakatua) and Elemenopi. We thank IUTCS, Brainstation23 and APSIS Solutions Ltd. for their effort and support towards ending this competition successfully. We are sharing the score-breakdown and some insights on the test data in the attached file. We will provide our detailed assessment of the solutions from different teams in a form of a research paper, soon. Stay tuned. </p>\n<p>We have a special prize for the researchers yet to be announced. We declare another prizemoney of 50k for the first team to achieve 30 WER on both the kaggle leaderboard and the hidden test set. If you are confident about your system, you need to claim for the prize by e-mailing the competition curators Asif Sushmit (sushmit@ieee.org) and Tahsin Reasat (greasat@gmail.com). The training and inference notebooks, model weights and data(if any) needs to be publicly shared in Kaggle to be eligible for the prizemoney. We are thankful to Apsis solutions ltd. for their generous contribution towards this. </p>\n<p>শেষ হইয়াও হইল না শেষ। Best of luck!</p>",
  "messages": [
    {
      "id": 2779211,
      "postDate": "2024-04-27T15:24:56.680Z",
      "content": "<p>Greetings from Bengali.AI</p>\n<p>We thank the research community for participating in the competition and present their approaches to tackle Bengali Regional Speech Recognition. We congratulate the top 3 teams:  Jibonanondo Das, কাকাতুয়া(kakatua) and Elemenopi. We thank IUTCS, Brainstation23 and APSIS Solutions Ltd. for their effort and support towards ending this competition successfully. We are sharing the score-breakdown and some insights on the test data in the attached file. We will provide our detailed assessment of the solutions from different teams in a form of a research paper, soon. Stay tuned. </p>\n<p>We have a special prize for the researchers yet to be announced. We declare another prizemoney of 50k for the first team to achieve 30 WER on both the kaggle leaderboard and the hidden test set. If you are confident about your system, you need to claim for the prize by e-mailing the competition curators Asif Sushmit (sushmit@ieee.org) and Tahsin Reasat (greasat@gmail.com). The training and inference notebooks, model weights and data(if any) needs to be publicly shared in Kaggle to be eligible for the prizemoney. We are thankful to Apsis solutions ltd. for their generous contribution towards this. </p>\n<p>শেষ হইয়াও হইল না শেষ। Best of luck!</p>",
      "rawMarkdown": "Greetings from Bengali.AI\n\nWe thank the research community for participating in the competition and present their approaches to tackle Bengali Regional Speech Recognition. We congratulate the top 3 teams:  Jibonanondo Das, কাকাতুয়া(kakatua) and Elemenopi. We thank IUTCS, Brainstation23 and APSIS Solutions Ltd. for their effort and support towards ending this competition successfully. We are sharing the score-breakdown and some insights on the test data in the attached file. We will provide our detailed assessment of the solutions from different teams in a form of a research paper, soon. Stay tuned. \n\nWe have a special prize for the researchers yet to be announced. We declare another prizemoney of 50k for the first team to achieve 30 WER on both the kaggle leaderboard and the hidden test set. If you are confident about your system, you need to claim for the prize by e-mailing the competition curators Asif Sushmit (sushmit@ieee.org) and Tahsin Reasat (greasat@gmail.com). The training and inference notebooks, model weights and data(if any) needs to be publicly shared in Kaggle to be eligible for the prizemoney. We are thankful to Apsis solutions ltd. for their generous contribution towards this. \n\nশেষ হইয়াও হইল না শেষ। Best of luck!",
      "votes": 3
    }
  ],
  "comments": [],
  "raw_markdown_by_id": {
    "2779211": "Greetings from Bengali.AI\n\nWe thank the research community for participating in the competition and present their approaches to tackle Bengali Regional Speech Recognition. We congratulate the top 3 teams:  Jibonanondo Das, কাকাতুয়া(kakatua) and Elemenopi. We thank IUTCS, Brainstation23 and APSIS Solutions Ltd. for their effort and support towards ending this competition successfully. We are sharing the score-breakdown and some insights on the test data in the attached file. We will provide our detailed assessment of the solutions from different teams in a form of a research paper, soon. Stay tuned. \n\nWe have a special prize for the researchers yet to be announced. We declare another prizemoney of 50k for the first team to achieve 30 WER on both the kaggle leaderboard and the hidden test set. If you are confident about your system, you need to claim for the prize by e-mailing the competition curators Asif Sushmit (sushmit@ieee.org) and Tahsin Reasat (greasat@gmail.com). The training and inference notebooks, model weights and data(if any) needs to be publicly shared in Kaggle to be eligible for the prizemoney. We are thankful to Apsis solutions ltd. for their generous contribution towards this. \n\nশেষ হইয়াও হইল না শেষ। Best of luck!"
  }
}