{ "id": 430989, "title": "Request to make your solution public - Competition Host", "url": "/competitions/dlsprint2/discussion/430989", "author_name": "Sameen53", "post_date": "2023-08-11T15:15:20.758000", "votes": 1, "comment_count": 0, "views": 0, "content": "

According to the competition rules, only the winners are required to make their solutions public.

\n

However, we would appreciate it if everyone makes their -

\n
    \n
  1. Code
    2. \n
  2. Model Weights
    3. \n
  3. Paper
    4. \n
  4. Presentation
    5. \n
\n

-public.

\n

Why you should make your solution public?

\n
    \n

  1. Document Layout Analysis is far from solved and the BaDLAD dataset goes a long way in contributing to this field. It is one of the largest datasets in any language and comes with 2M+ unannotated data as well. However, no dataset is complete without a community actively utilizing it. Your codes, model weights, papers and presentations will be an invaluable resource to anyone who will work on the dataset next. They will already know what worked and what didn't, and can immediately start working on improving the state-of-the-art.

    2. \n

  2. Everyone will have access to your resources and you will have access to theirs. This is a competition but ultimately, not a zero-sum game. We encourage the participants to turn their work into a publication, it doesn't need to be the winning solution, you could write an analysis paper, meta-report etc. Bengali.AI and BUET CSE'18 will fully support this endeavour.

    3. \n

  3. We will email a Finalist Certificate to teams that make their solution public. 🤗

    4. \n

  4. Bengali.AI and BUET CSE'18 plan to make a competition report paper. So we will cite all finalist papers. Please upload your paper to arxiv to allow us to cite it.

    5. \n
\n

How to Do It:

\n
    \n
  1. Make your code public.
    2. \n
  2. Make your dataset public.
    3. \n
  3. Submit your paper to arxiv. Mail us if you need help with this.
    4. \n
  4. Create a discussion with the title Team-TeamName-Solution.
    5. \n
  5. Link to Template Solution.
    6. \n
  6. Check the checkbox - This is my team's solution write-up.
    7. \n
\n

Thanks.

", "messages": [ { "id": 2385827, "postDate": "2023-08-11T15:15:20.760Z", "content": "

According to the competition rules, only the winners are required to make their solutions public.

\n

However, we would appreciate it if everyone makes their -

\n
    \n
  1. Code
    2. \n
  2. Model Weights
    3. \n
  3. Paper
    4. \n
  4. Presentation
    5. \n
\n

-public.

\n

Why you should make your solution public?

\n
    \n

  1. Document Layout Analysis is far from solved and the BaDLAD dataset goes a long way in contributing to this field. It is one of the largest datasets in any language and comes with 2M+ unannotated data as well. However, no dataset is complete without a community actively utilizing it. Your codes, model weights, papers and presentations will be an invaluable resource to anyone who will work on the dataset next. They will already know what worked and what didn't, and can immediately start working on improving the state-of-the-art.

    2. \n

  2. Everyone will have access to your resources and you will have access to theirs. This is a competition but ultimately, not a zero-sum game. We encourage the participants to turn their work into a publication, it doesn't need to be the winning solution, you could write an analysis paper, meta-report etc. Bengali.AI and BUET CSE'18 will fully support this endeavour.

    3. \n

  3. We will email a Finalist Certificate to teams that make their solution public. 🤗

    4. \n

  4. Bengali.AI and BUET CSE'18 plan to make a competition report paper. So we will cite all finalist papers. Please upload your paper to arxiv to allow us to cite it.

    5. \n
\n

How to Do It:

\n
    \n
  1. Make your code public.
    2. \n
  2. Make your dataset public.
    3. \n
  3. Submit your paper to arxiv. Mail us if you need help with this.
    4. \n
  4. Create a discussion with the title Team-TeamName-Solution.
    5. \n
  5. Link to Template Solution.
    6. \n
  6. Check the checkbox - This is my team's solution write-up.
    7. \n
\n

Thanks.

", "rawMarkdown": "According to the competition rules, **only the winners are required to make their solutions public.**\n\nHowever, we would appreciate it if everyone makes their - \n\n1. Code\n2. Model Weights\n3. Paper\n4. Presentation\n\n-public.\n\n### Why you should make your solution public?\n1. Document Layout Analysis is far from solved and the BaDLAD dataset goes a long way in contributing to this field. It is one of the largest datasets in any language and comes with 2M+ unannotated data as well. However, no dataset is complete without a community actively utilizing it. Your codes, model weights, papers and presentations will be an invaluable resource to anyone who will work on the dataset next. They will already know what worked and what didn't, and can immediately start working on improving the state-of-the-art.\n\n2. Everyone will have access to your resources and you will have access to theirs. This is a competition but ultimately, not a zero-sum game. We encourage the participants to turn their work into a publication, it doesn't need to be the winning solution, you could write an analysis paper, meta-report etc. Bengali.AI and BUET CSE'18 will fully support this endeavour.\n\n3. We will email a Finalist Certificate to teams that make their solution public. 🤗 \n\n4. Bengali.AI and BUET CSE'18 plan to make a competition report paper. So we will cite all finalist papers. Please upload your paper to [arxiv](https://arxiv.org/) to allow us to cite it. \n\n### How to Do It:\n1. Make your code public.\n2. Make your dataset public.\n3. Submit your paper to [arxiv](https://arxiv.org/). Mail us if you need help with this.\n4. Create a discussion with the title `Team-TeamName-Solution`.\n5. [Link to Template Solution](https://www.kaggle.com/competitions/dlsprint2/discussion/430985).\n6. Check the checkbox - **This is my team's solution write-up**.\n\nThanks.", "votes": 1 } ], "comments": [], "raw_markdown_by_id": { "2385827": "According to the competition rules, **only the winners are required to make their solutions public.**\n\nHowever, we would appreciate it if everyone makes their - \n\n1. Code\n2. Model Weights\n3. Paper\n4. Presentation\n\n-public.\n\n### Why you should make your solution public?\n1. Document Layout Analysis is far from solved and the BaDLAD dataset goes a long way in contributing to this field. It is one of the largest datasets in any language and comes with 2M+ unannotated data as well. However, no dataset is complete without a community actively utilizing it. Your codes, model weights, papers and presentations will be an invaluable resource to anyone who will work on the dataset next. They will already know what worked and what didn't, and can immediately start working on improving the state-of-the-art.\n\n2. Everyone will have access to your resources and you will have access to theirs. This is a competition but ultimately, not a zero-sum game. We encourage the participants to turn their work into a publication, it doesn't need to be the winning solution, you could write an analysis paper, meta-report etc. Bengali.AI and BUET CSE'18 will fully support this endeavour.\n\n3. We will email a Finalist Certificate to teams that make their solution public. 🤗 \n\n4. Bengali.AI and BUET CSE'18 plan to make a competition report paper. So we will cite all finalist papers. Please upload your paper to [arxiv](https://arxiv.org/) to allow us to cite it. \n\n### How to Do It:\n1. Make your code public.\n2. Make your dataset public.\n3. Submit your paper to [arxiv](https://arxiv.org/). Mail us if you need help with this.\n4. Create a discussion with the title `Team-TeamName-Solution`.\n5. [Link to Template Solution](https://www.kaggle.com/competitions/dlsprint2/discussion/430985).\n6. Check the checkbox - **This is my team's solution write-up**.\n\nThanks." } }