Hi fellow teams!
\n\nThanks so much for your participation during the last 2 months!
\n\nSince the end of the competition is approaching, I would like to collect information about who is going to CVPR this year. If you have already registered, please let me know if you are willing to attend our workshop and/or are willing to share your insights (format could be oral/poster, TBD).
\n\nIf your answer to the questions above is yes, please directly contact me to RSVP. Or you can reply to this thread and I will contact you.
\n\nLooking forward to your final submissions and meeting you at the workshop!
", "messages": [ { "id": 538609, "postDate": "2019-05-28T20:59:33.130Z", "content": "Hi fellow teams!
\n\nThanks so much for your participation during the last 2 months!
\n\nSince the end of the competition is approaching, I would like to collect information about who is going to CVPR this year. If you have already registered, please let me know if you are willing to attend our workshop and/or are willing to share your insights (format could be oral/poster, TBD).
\n\nIf your answer to the questions above is yes, please directly contact me to RSVP. Or you can reply to this thread and I will contact you.
\n\nLooking forward to your final submissions and meeting you at the workshop!
", "rawMarkdown": "Hi fellow teams!\n\nThanks so much for your participation during the last 2 months!\n\nSince the end of the competition is approaching, I would like to collect information about who is going to CVPR this year. If you have already registered, please let me know if you are willing to attend our [workshop](https://sites.google.com/corp/view/fgvc6/hom) and/or are willing to share your insights (format could be oral/poster, TBD).\n\nIf your answer to the questions above is yes, please directly contact me to RSVP. Or you can reply to this thread and I will contact you.\n\nLooking forward to your final submissions and meeting you at the workshop!", "votes": 4 }, { "id": 543746, "postDate": "2019-06-04T19:00:06.187Z", "content": "Since attendees need to pay all costs associated with travel, expenses, and fees by themselves. I don't think there will be a lot of competitors going to CVPR. The costs at this time to go to LA is too expensive. No idea why no sponsor for these costs
", "rawMarkdown": "Since attendees need to pay all costs associated with travel, expenses, and fees by themselves. I don't think there will be a lot of competitors going to CVPR. The costs at this time to go to LA is too expensive. No idea why no sponsor for these costs", "votes": 1, "replies": [ { "id": 544764, "postDate": "2019-06-05T22:14:05.267Z", "content": "because image classification is not \"hot topic\" now.
", "rawMarkdown": "because image classification is not \"hot topic\" now." }, { "id": 549502, "postDate": "2019-06-10T18:12:36.723Z", "content": "Hey @strideradu, is anyone from your team is already going to CVPR? Thanks.
", "rawMarkdown": "Hey @strideradu, is anyone from your team is already going to CVPR? Thanks." }, { "id": 549570, "postDate": "2019-06-10T20:12:21.810Z", "content": "I don't think so, since most of my teammates are in China and Japan now, and I don't think they have time to get a visa. Thanks for asking!
", "rawMarkdown": "I don't think so, since most of my teammates are in China and Japan now, and I don't think they have time to get a visa. Thanks for asking!" } ] }, { "id": 549332, "postDate": "2019-06-10T15:01:18.537Z", "rawMarkdown": "", "votes": 1, "isDeleted": true, "replies": [ { "id": 549338, "postDate": "2019-06-10T15:05:13.477Z", "content": "@wcukierski Is there any quick fix that we can apply?
", "rawMarkdown": "@wcukierski Is there any quick fix that we can apply?" }, { "id": 549340, "postDate": "2019-06-10T15:12:48.813Z", "content": "is that fair to other players?
", "rawMarkdown": "is that fair to other players?", "votes": 1 }, { "id": 549342, "postDate": "2019-06-10T15:15:23.297Z", "content": "does other player have a change to modify the kernel if their kernel failed in second stage ?
", "rawMarkdown": "does other player have a change to modify the kernel if their kernel failed in second stage ?" }, { "id": 549345, "postDate": "2019-06-10T15:20:20.850Z", "content": "Sorry, we just want to modify a small program bug 、a number, we did not modify the algorithm model, algorithm implementation and program logic. The kernel only game is just beginning, we are not familiar with it, and there is not enough experience to test.
", "rawMarkdown": "Sorry, we just want to modify a small program bug 、a number, we did not modify the algorithm model, algorithm implementation and program logic. The kernel only game is just beginning, we are not familiar with it, and there is not enough experience to test." }, { "id": 549346, "postDate": "2019-06-10T15:20:40.083Z", "content": "thanks for your reply
", "rawMarkdown": "thanks for your reply" }, { "id": 549349, "postDate": "2019-06-10T15:22:30.393Z", "content": "We just want to fix bugs in the program, not change the algorithm. We've been trying to improve the algorithm for two months, and it's an algorithmic contest.\nWe really want to share our innovative approach after the contest. Based on the algorithm in paper “Improving CNN Classifiers by Estimating Test-time Priors”, we took advantage of the single-label model of the original paper and applied estimated test-time priors to a multi-label model of. However, it will be a pity if we cannot share it after all these explorations and efforts due to a small mistake in the program. so we really hope that we can have a chance to get a true score that can truly reflect our efforts.
", "rawMarkdown": "We just want to fix bugs in the program, not change the algorithm. We've been trying to improve the algorithm for two months, and it's an algorithmic contest.\nWe really want to share our innovative approach after the contest. Based on the algorithm in paper “Improving CNN Classifiers by Estimating Test-time Priors”, we took advantage of the single-label model of the original paper and applied estimated test-time priors to a multi-label model of. However, it will be a pity if we cannot share it after all these explorations and efforts due to a small mistake in the program. so we really hope that we can have a chance to get a true score that can truly reflect our efforts.", "votes": 1 }, { "id": 549362, "postDate": "2019-06-10T15:36:53.777Z", "content": "I am Ok if kaggle allows you to modify , but at least in that case , other player should have the same right as you.
", "rawMarkdown": "I am Ok if kaggle allows you to modify , but at least in that case , other player should have the same right as you." }, { "id": 549373, "postDate": "2019-06-10T15:41:04.063Z", "content": "thank you . I think kaggle's kernel only rule has just begun and should be perfected in the future.
", "rawMarkdown": "thank you . I think kaggle's kernel only rule has just begun and should be perfected in the future. " }, { "id": 549412, "postDate": "2019-06-10T16:08:21.953Z", "content": "So , does kaggle allow you to fix or not ?
", "rawMarkdown": "So , does kaggle allow you to fix or not ?" }, { "id": 549425, "postDate": "2019-06-10T16:25:01.917Z", "content": "kaggle office give us a email don't allow this kind operation. they are afraid of infinite delay
", "rawMarkdown": "kaggle office give us a email don't allow this kind operation. they are afraid of infinite delay", "votes": -1 }, { "id": 549442, "postDate": "2019-06-10T16:40:30.333Z", "content": "Hi Team hy-cn.bj.xa, at this time we cannot support your request due to the limit of time (we are already delayed for a week and cannot afford allowing teams do changes at this scale). However, please let me know if you are still willing to share your tech to the workshop if you are going to CVPR. I am willing to discuss with you about details. Thanks for understanding!
", "rawMarkdown": "Hi Team hy-cn.bj.xa, at this time we cannot support your request due to the limit of time (we are already delayed for a week and cannot afford allowing teams do changes at this scale). However, please let me know if you are still willing to share your tech to the workshop if you are going to CVPR. I am willing to discuss with you about details. Thanks for understanding!" }, { "id": 549446, "postDate": "2019-06-10T16:47:44.130Z", "content": "OK,thank you ,we will share our tech later
", "rawMarkdown": "OK,thank you ,we will share our tech later" } ] }, { "id": 549330, "postDate": "2019-06-10T14:58:50.067Z", "rawMarkdown": "", "isDeleted": true }, { "id": 543863, "postDate": "2019-06-04T22:29:01.977Z", "rawMarkdown": "", "isDeleted": true } ], "comments": [ { "id": 543746, "author_name": "Strideradu", "author_url": "", "post_date": "2019-06-04T19:00:06.187000", "content": "Since attendees need to pay all costs associated with travel, expenses, and fees by themselves. I don't think there will be a lot of competitors going to CVPR. The costs at this time to go to LA is too expensive. No idea why no sponsor for these costs
", "votes": 1, "replies": [ { "id": 544764, "author_name": "soywu", "author_url": "", "post_date": "2019-06-05T22:14:05.267000", "content": "because image classification is not \"hot topic\" now.
", "votes": 0, "replies": [] }, { "id": 549502, "author_name": "Chenyang Zhang", "author_url": "", "post_date": "2019-06-10T18:12:36.723000", "content": "Hey @strideradu, is anyone from your team is already going to CVPR? Thanks.
", "votes": 0, "replies": [] }, { "id": 549570, "author_name": "Strideradu", "author_url": "", "post_date": "2019-06-10T20:12:21.810000", "content": "I don't think so, since most of my teammates are in China and Japan now, and I don't think they have time to get a visa. Thanks for asking!
", "votes": 0, "replies": [] } ] }, { "id": 549332, "author_name": "", "author_url": "", "post_date": "2019-06-10T15:01:18.537000", "content": "", "votes": 1, "replies": [ { "id": 549338, "author_name": "Chenyang Zhang", "author_url": "", "post_date": "2019-06-10T15:05:13.477000", "content": "@wcukierski Is there any quick fix that we can apply?
", "votes": 0, "replies": [] }, { "id": 549340, "author_name": "soywu", "author_url": "", "post_date": "2019-06-10T15:12:48.813000", "content": "is that fair to other players?
", "votes": 1, "replies": [] }, { "id": 549342, "author_name": "soywu", "author_url": "", "post_date": "2019-06-10T15:15:23.297000", "content": "does other player have a change to modify the kernel if their kernel failed in second stage ?
", "votes": 0, "replies": [] }, { "id": 549345, "author_name": "", "author_url": "", "post_date": "2019-06-10T15:20:20.850000", "content": "Sorry, we just want to modify a small program bug 、a number, we did not modify the algorithm model, algorithm implementation and program logic. The kernel only game is just beginning, we are not familiar with it, and there is not enough experience to test.
", "votes": 0, "replies": [] }, { "id": 549346, "author_name": "", "author_url": "", "post_date": "2019-06-10T15:20:40.083000", "content": "thanks for your reply
", "votes": 0, "replies": [] }, { "id": 549349, "author_name": "", "author_url": "", "post_date": "2019-06-10T15:22:30.393000", "content": "We just want to fix bugs in the program, not change the algorithm. We've been trying to improve the algorithm for two months, and it's an algorithmic contest.\nWe really want to share our innovative approach after the contest. Based on the algorithm in paper “Improving CNN Classifiers by Estimating Test-time Priors”, we took advantage of the single-label model of the original paper and applied estimated test-time priors to a multi-label model of. However, it will be a pity if we cannot share it after all these explorations and efforts due to a small mistake in the program. so we really hope that we can have a chance to get a true score that can truly reflect our efforts.
", "votes": 1, "replies": [] }, { "id": 549362, "author_name": "soywu", "author_url": "", "post_date": "2019-06-10T15:36:53.777000", "content": "I am Ok if kaggle allows you to modify , but at least in that case , other player should have the same right as you.
", "votes": 0, "replies": [] }, { "id": 549373, "author_name": "", "author_url": "", "post_date": "2019-06-10T15:41:04.063000", "content": "thank you . I think kaggle's kernel only rule has just begun and should be perfected in the future.
", "votes": 0, "replies": [] }, { "id": 549412, "author_name": "soywu", "author_url": "", "post_date": "2019-06-10T16:08:21.953000", "content": "So , does kaggle allow you to fix or not ?
", "votes": 0, "replies": [] }, { "id": 549425, "author_name": "", "author_url": "", "post_date": "2019-06-10T16:25:01.917000", "content": "kaggle office give us a email don't allow this kind operation. they are afraid of infinite delay
", "votes": -1, "replies": [] }, { "id": 549442, "author_name": "Chenyang Zhang", "author_url": "", "post_date": "2019-06-10T16:40:30.333000", "content": "Hi Team hy-cn.bj.xa, at this time we cannot support your request due to the limit of time (we are already delayed for a week and cannot afford allowing teams do changes at this scale). However, please let me know if you are still willing to share your tech to the workshop if you are going to CVPR. I am willing to discuss with you about details. Thanks for understanding!
", "votes": 0, "replies": [] }, { "id": 549446, "author_name": "", "author_url": "", "post_date": "2019-06-10T16:47:44.130000", "content": "OK,thank you ,we will share our tech later
", "votes": 0, "replies": [] } ] }, { "id": 549330, "author_name": "", "author_url": "", "post_date": "2019-06-10T14:58:50.067000", "content": "", "votes": 0, "replies": [] }, { "id": 543863, "author_name": "", "author_url": "", "post_date": "2019-06-04T22:29:01.977000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "538609": "Hi fellow teams!\n\nThanks so much for your participation during the last 2 months!\n\nSince the end of the competition is approaching, I would like to collect information about who is going to CVPR this year. If you have already registered, please let me know if you are willing to attend our [workshop](https://sites.google.com/corp/view/fgvc6/hom) and/or are willing to share your insights (format could be oral/poster, TBD).\n\nIf your answer to the questions above is yes, please directly contact me to RSVP. Or you can reply to this thread and I will contact you.\n\nLooking forward to your final submissions and meeting you at the workshop!", "543746": "Since attendees need to pay all costs associated with travel, expenses, and fees by themselves. I don't think there will be a lot of competitors going to CVPR. The costs at this time to go to LA is too expensive. No idea why no sponsor for these costs", "549332": "", "549330": "", "543863": "" } }