{ "id": 171715, "title": "Kind Reminder for Submission Compliance", "url": "/competitions/global-wheat-detection/discussion/171715", "author_name": "", "post_date": "2020-08-02T07:11:40.980081Z", "votes": 45, "comment_count": 16, "views": 0, "content": "

As the deadline approaches, I would like to kindly remind all contestants to carefully select your final submissions.

\n\n

Though Kaggle is a great community for studying and supporting each other, a crucial part remains to be hosting competitions, and this challenge, in particular, is also a competition. This means that those accepting the rules should obey them carefully. And unlike a submission-based competition, where its sort of \"easier\" to cheat by manually overriding the csv, the organizers will see what you wrote exactly for your submission, line by line.

\n\n

Please don't select non-compliant submissions as any of your final submissions. It is very tempting to do so, but please don't just think that \"If I can do compliant and non-compliant I can still get a decent ranking even if the latter is banned.\" You would be wasting everyone's time and effort. Organizers will eventually find out and you will be removed from the leaderboard, not to mention losing the respects of your fellow Kagglers for cheating.

\n\n

There will be more detection-based challenges in the future, and you don't need this one to be the definitive factor in your rise on the global ranking, where a non-compliant model just happens to be well-performing.

\n\n

There is a certain weight to a top ranking in Kaggle challenges as companies are taking into consideration of your placements when hiring people, so let's make a collaborative effort together to keep it that way.

\n\n

Sincerely,\nAlex

", "messages": [ { "id": "954890", "postDate": "08/02/2020 07:11:40", "content": "

As the deadline approaches, I would like to kindly remind all contestants to carefully select your final submissions.

\n\n

Though Kaggle is a great community for studying and supporting each other, a crucial part remains to be hosting competitions, and this challenge, in particular, is also a competition. This means that those accepting the rules should obey them carefully. And unlike a submission-based competition, where its sort of \"easier\" to cheat by manually overriding the csv, the organizers will see what you wrote exactly for your submission, line by line.

\n\n

Please don't select non-compliant submissions as any of your final submissions. It is very tempting to do so, but please don't just think that \"If I can do compliant and non-compliant I can still get a decent ranking even if the latter is banned.\" You would be wasting everyone's time and effort. Organizers will eventually find out and you will be removed from the leaderboard, not to mention losing the respects of your fellow Kagglers for cheating.

\n\n

There will be more detection-based challenges in the future, and you don't need this one to be the definitive factor in your rise on the global ranking, where a non-compliant model just happens to be well-performing.

\n\n

There is a certain weight to a top ranking in Kaggle challenges as companies are taking into consideration of your placements when hiring people, so let's make a collaborative effort together to keep it that way.

\n\n

Sincerely,\nAlex

", "rawMarkdown": "As the deadline approaches, I would like to kindly remind all contestants to carefully select your final submissions.\n\nThough Kaggle is a great community for studying and supporting each other, a crucial part remains to be hosting competitions, and this challenge, in particular, is also a competition. This means that those accepting the rules should obey them carefully. And unlike a submission-based competition, where its sort of \"easier\" to cheat by manually overriding the csv, the organizers will see what you wrote exactly for your submission, line by line.\n\nPlease don't select non-compliant submissions as any of your final submissions. It is very tempting to do so, but please don't just think that \"If I can do compliant and non-compliant I can still get a decent ranking even if the latter is banned.\" You would be wasting everyone's time and effort. Organizers will eventually find out and you will be removed from the leaderboard, not to mention losing the respects of your fellow Kagglers for cheating. \n\nThere will be more detection-based challenges in the future, and you don't need this one to be the definitive factor in your rise on the global ranking, where a non-compliant model just happens to be well-performing. \n\nThere is a certain weight to a top ranking in Kaggle challenges as companies are taking into consideration of your placements when hiring people, so let's make a collaborative effort together to keep it that way.\n\nSincerely,\nAlex", "votes": null }, { "id": "955109", "postDate": "08/02/2020 10:53:45", "content": "

Will you share your solutions after this competition, I think I can learn a lot from it๐Ÿ˜„

", "rawMarkdown": "Will you share your solutions after this competition, I think I can learn a lot from it๐Ÿ˜„", "votes": null }, { "id": "955561", "postDate": "08/02/2020 17:43:44", "content": "

Well said Alex. After the competition ends will be interesting to read anything you are willing to share about your solution.

", "rawMarkdown": "Well said Alex. After the competition ends will be interesting to read anything you are willing to share about your solution.", "votes": null }, { "id": "955856", "postDate": "08/03/2020 02:29:16", "content": "

I think someone will sneak to use yolov5 as final submit.

", "rawMarkdown": "I think someone will sneak to use yolov5 as final submit.", "votes": null }, { "id": "955891", "postDate": "08/03/2020 04:01:27", "content": "

The leaderboard is so messed up with all this YOLOv5 situation. I wonder how kaggle admins will handle it.

", "rawMarkdown": "The leaderboard is so messed up with all this YOLOv5 situation. I wonder how kaggle admins will handle it.", "votes": null }, { "id": "956015", "postDate": "08/03/2020 06:49:25", "content": "

I think we should expect a huge shake-up just cause two following reasons:\n1) A lot of broken boxes in train, public and private datasets. I'm sure that a lot of people, in particular even those guys who are in the medal zone now, don't know this fact. And all it cost to read the necessary discussion.\n2) A lot of people using Yolov5, and again in particular in the medal zone. I'm talking exactly about their the best scores, they got it by Yolov5. A part of them will just choose another submit and apparently will drop down, another part of them will state in the final submission Yolov5 solution and will definitely drop down in the private LB.

", "rawMarkdown": "I think we should expect a huge shake-up just cause two following reasons:\n1) A lot of broken boxes in train, public and private datasets. I'm sure that a lot of people, in particular even those guys who are in the medal zone now, don't know this fact. And all it cost to read the necessary discussion.\n2) A lot of people using Yolov5, and again in particular in the medal zone. I'm talking exactly about their the best scores, they got it by Yolov5. A part of them will just choose another submit and apparently will drop down, another part of them will state in the final submission Yolov5 solution and will definitely drop down in the private LB.", "votes": null }, { "id": "956275", "postDate": "08/03/2020 11:38:44", "content": "

I guess some people didn't notice that yolov5 can't be used..

", "rawMarkdown": "I guess some people didn't notice that yolov5 can't be used..", "votes": null }, { "id": "956349", "postDate": "08/03/2020 12:34:03", "content": "

If they use it what will happen? I will not use yolov5

", "rawMarkdown": "If they use it what will happen? I will not use yolov5", "votes": null }, { "id": "956356", "postDate": "08/03/2020 12:40:37", "content": "

Maybe Kaggle host will check kernel. Then yolov5 competitor is removed Leaderboard if this is correct

", "rawMarkdown": "Maybe Kaggle host will check kernel. Then yolov5 competitor is removed Leaderboard if this is correct", "votes": null }, { "id": "957559", "postDate": "08/04/2020 12:06:41", "content": "

Many will sneak to use yolov5 as final submission

", "rawMarkdown": "Many will sneak to use yolov5 as final submission", "votes": null }, { "id": "957607", "postDate": "08/04/2020 12:46:07", "content": "

@tereka That is the ideal case. Hope they check entire medal zone instead of just in the money zone and remove all non-compliant solutions. Only this will ensure fair results.

", "rawMarkdown": "tereka That is the ideal case. Hope they check entire medal zone instead of just in the money zone and remove all non-compliant solutions. Only this will ensure fair results.", "votes": null }, { "id": "957624", "postDate": "08/04/2020 12:57:34", "content": "

I agree. In past competition, Kaggle host checked all kernels for fair results.\nhttps://www.kaggle.com/c/generative-dog-images/discussion/106206

", "rawMarkdown": "I agree. In past competition, Kaggle host checked all kernels for fair results.\nhttps://www.kaggle.com/c/generative-dog-images/discussion/106206", "votes": null }, { "id": "957714", "postDate": "08/04/2020 14:04:58", "content": "

@tereka Oh, interesting! I didn't think Kaggle can manually check submissions.

", "rawMarkdown": "tereka Oh, interesting! I didn't think Kaggle can manually check submissions.", "votes": null }, { "id": "957770", "postDate": "08/04/2020 14:51:14", "content": "

Manually check is very difficult. because this comp is bigger than past comp. \nMaybe kaggle use manually checking and automatically tool.

", "rawMarkdown": "Manually check is very difficult. because this comp is bigger than past comp. \nMaybe kaggle use manually checking and automatically tool.", "votes": null }, { "id": "957816", "postDate": "08/04/2020 15:16:16", "content": "

I hope Kaggle could eventually invalidate all non-compliant solutions. IMHO, It would be really discouraging if the rule makers cannot guarantee the interests of the people who follow them.

", "rawMarkdown": "I hope Kaggle could eventually invalidate all non-compliant solutions. IMHO, It would be really discouraging if the rule makers cannot guarantee the interests of the people who follow them.", "votes": null }, { "id": "957818", "postDate": "08/04/2020 15:18:32", "content": "

@tereka \nGood find on the past competition. They took 15 days in Generative Dog Images competition to verify all submissions, but they actually DID IT !

", "rawMarkdown": "tereka \nGood find on the past competition. They took 15 days in Generative Dog Images competition to verify all submissions, but they actually DID IT !", "votes": null }, { "id": "957918", "postDate": "08/04/2020 16:19:13", "content": "

Nice

", "rawMarkdown": "Nice", "votes": null } ], "comments": [ { "id": 955109, "author_name": "m1nglei", "author_url": "", "post_date": "08/02/2020 10:53:45", "content": "

Will you share your solutions after this competition, I think I can learn a lot from it๐Ÿ˜„

", "votes": null, "replies": [] }, { "id": 955561, "author_name": "datahobbit", "author_url": "", "post_date": "08/02/2020 17:43:44", "content": "

Well said Alex. After the competition ends will be interesting to read anything you are willing to share about your solution.

", "votes": null, "replies": [] }, { "id": 955856, "author_name": "zekunn", "author_url": "", "post_date": "08/03/2020 02:29:16", "content": "

I think someone will sneak to use yolov5 as final submit.

", "votes": null, "replies": [ { "id": 957559, "author_name": "rifat963", "author_url": "", "post_date": "08/04/2020 12:06:41", "content": "

Many will sneak to use yolov5 as final submission

", "votes": null, "replies": [] } ] }, { "id": 955891, "author_name": "ahmmadrashid", "author_url": "", "post_date": "08/03/2020 04:01:27", "content": "

The leaderboard is so messed up with all this YOLOv5 situation. I wonder how kaggle admins will handle it.

", "votes": null, "replies": [] }, { "id": 956015, "author_name": "vadimtimakin", "author_url": "", "post_date": "08/03/2020 06:49:25", "content": "

I think we should expect a huge shake-up just cause two following reasons:\n1) A lot of broken boxes in train, public and private datasets. I'm sure that a lot of people, in particular even those guys who are in the medal zone now, don't know this fact. And all it cost to read the necessary discussion.\n2) A lot of people using Yolov5, and again in particular in the medal zone. I'm talking exactly about their the best scores, they got it by Yolov5. A part of them will just choose another submit and apparently will drop down, another part of them will state in the final submission Yolov5 solution and will definitely drop down in the private LB.

", "votes": null, "replies": [] }, { "id": 956275, "author_name": "tereka", "author_url": "", "post_date": "08/03/2020 11:38:44", "content": "

I guess some people didn't notice that yolov5 can't be used..

", "votes": null, "replies": [ { "id": 956349, "author_name": "zekunn", "author_url": "", "post_date": "08/03/2020 12:34:03", "content": "

If they use it what will happen? I will not use yolov5

", "votes": null, "replies": [] }, { "id": 956356, "author_name": "tereka", "author_url": "", "post_date": "08/03/2020 12:40:37", "content": "

Maybe Kaggle host will check kernel. Then yolov5 competitor is removed Leaderboard if this is correct

", "votes": null, "replies": [] }, { "id": 957607, "author_name": "kaushal2896", "author_url": "", "post_date": "08/04/2020 12:46:07", "content": "

@tereka That is the ideal case. Hope they check entire medal zone instead of just in the money zone and remove all non-compliant solutions. Only this will ensure fair results.

", "votes": null, "replies": [] }, { "id": 957624, "author_name": "tereka", "author_url": "", "post_date": "08/04/2020 12:57:34", "content": "

I agree. In past competition, Kaggle host checked all kernels for fair results.\nhttps://www.kaggle.com/c/generative-dog-images/discussion/106206

", "votes": null, "replies": [] }, { "id": 957714, "author_name": "sergeyzlobin", "author_url": "", "post_date": "08/04/2020 14:04:58", "content": "

@tereka Oh, interesting! I didn't think Kaggle can manually check submissions.

", "votes": null, "replies": [] }, { "id": 957770, "author_name": "tereka", "author_url": "", "post_date": "08/04/2020 14:51:14", "content": "

Manually check is very difficult. because this comp is bigger than past comp. \nMaybe kaggle use manually checking and automatically tool.

", "votes": null, "replies": [] }, { "id": 957818, "author_name": "watzisname", "author_url": "", "post_date": "08/04/2020 15:18:32", "content": "

@tereka \nGood find on the past competition. They took 15 days in Generative Dog Images competition to verify all submissions, but they actually DID IT !

", "votes": null, "replies": [] } ] }, { "id": 957816, "author_name": "zacchaeus", "author_url": "", "post_date": "08/04/2020 15:16:16", "content": "

I hope Kaggle could eventually invalidate all non-compliant solutions. IMHO, It would be really discouraging if the rule makers cannot guarantee the interests of the people who follow them.

", "votes": null, "replies": [] }, { "id": 957918, "author_name": "drskfuncuts", "author_url": "", "post_date": "08/04/2020 16:19:13", "content": "

Nice

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "954890": "As the deadline approaches, I would like to kindly remind all contestants to carefully select your final submissions.\n\nThough Kaggle is a great community for studying and supporting each other, a crucial part remains to be hosting competitions, and this challenge, in particular, is also a competition. This means that those accepting the rules should obey them carefully. And unlike a submission-based competition, where its sort of \"easier\" to cheat by manually overriding the csv, the organizers will see what you wrote exactly for your submission, line by line.\n\nPlease don't select non-compliant submissions as any of your final submissions. It is very tempting to do so, but please don't just think that \"If I can do compliant and non-compliant I can still get a decent ranking even if the latter is banned.\" You would be wasting everyone's time and effort. Organizers will eventually find out and you will be removed from the leaderboard, not to mention losing the respects of your fellow Kagglers for cheating. \n\nThere will be more detection-based challenges in the future, and you don't need this one to be the definitive factor in your rise on the global ranking, where a non-compliant model just happens to be well-performing. \n\nThere is a certain weight to a top ranking in Kaggle challenges as companies are taking into consideration of your placements when hiring people, so let's make a collaborative effort together to keep it that way.\n\nSincerely,\nAlex", "955109": "Will you share your solutions after this competition, I think I can learn a lot from it๐Ÿ˜„", "955561": "Well said Alex. After the competition ends will be interesting to read anything you are willing to share about your solution.", "955856": "I think someone will sneak to use yolov5 as final submit.", "955891": "The leaderboard is so messed up with all this YOLOv5 situation. I wonder how kaggle admins will handle it.", "956015": "I think we should expect a huge shake-up just cause two following reasons:\n1) A lot of broken boxes in train, public and private datasets. I'm sure that a lot of people, in particular even those guys who are in the medal zone now, don't know this fact. And all it cost to read the necessary discussion.\n2) A lot of people using Yolov5, and again in particular in the medal zone. I'm talking exactly about their the best scores, they got it by Yolov5. A part of them will just choose another submit and apparently will drop down, another part of them will state in the final submission Yolov5 solution and will definitely drop down in the private LB.", "956275": "I guess some people didn't notice that yolov5 can't be used..", "956349": "If they use it what will happen? I will not use yolov5", "956356": "Maybe Kaggle host will check kernel. Then yolov5 competitor is removed Leaderboard if this is correct", "957559": "Many will sneak to use yolov5 as final submission", "957607": "tereka That is the ideal case. Hope they check entire medal zone instead of just in the money zone and remove all non-compliant solutions. Only this will ensure fair results.", "957624": "I agree. In past competition, Kaggle host checked all kernels for fair results.\nhttps://www.kaggle.com/c/generative-dog-images/discussion/106206", "957714": "tereka Oh, interesting! I didn't think Kaggle can manually check submissions.", "957770": "Manually check is very difficult. because this comp is bigger than past comp. \nMaybe kaggle use manually checking and automatically tool.", "957816": "I hope Kaggle could eventually invalidate all non-compliant solutions. IMHO, It would be really discouraging if the rule makers cannot guarantee the interests of the people who follow them.", "957818": "tereka \nGood find on the past competition. They took 15 days in Generative Dog Images competition to verify all submissions, but they actually DID IT !", "957918": "Nice" }, "source": "meta" }