Dear Kagglers,
\n\nOn behalf of Airbus DS Intelligence and our R&D team, I would like to officially welcome you to the “Airbus Ship Detection Challenge”!
\n\nThis competition is about finding ships on satellite images as quickly as possible. To fight against potentially devastating events at sea like ship accidents, piracy or illegal fishing, we need to act fast. This is why we challenge you to help us detect ALL ships on our satellite images with machine learning and do it FAST enough that we can send the alert on time.
\n\nWe have provided a large quantity of labeled images from our constellation of SPOT satellites as well as an even bigger collection of images of sea without ships. This is because finding ships in high seas is like finding a needle in a haystack! I am sure that this will generate a lot of innovation on models and on training. But do not forget about the Algorithm Speed Prize! Speed is very valuable as well. Please feel free to ask us any questions regarding the data, the competition, our R&D team or our company.
\n\nWe hope you will have a lot of fun with this competition!\nLooking forward to hearing from you in the discussion!
\n\nThe R&D team from Airbus DS Intelligence.
", "messages": [ { "id": 364140, "postDate": "2018-07-30T20:47:05.043Z", "content": "Dear Kagglers,
\n\nOn behalf of Airbus DS Intelligence and our R&D team, I would like to officially welcome you to the “Airbus Ship Detection Challenge”!
\n\nThis competition is about finding ships on satellite images as quickly as possible. To fight against potentially devastating events at sea like ship accidents, piracy or illegal fishing, we need to act fast. This is why we challenge you to help us detect ALL ships on our satellite images with machine learning and do it FAST enough that we can send the alert on time.
\n\nWe have provided a large quantity of labeled images from our constellation of SPOT satellites as well as an even bigger collection of images of sea without ships. This is because finding ships in high seas is like finding a needle in a haystack! I am sure that this will generate a lot of innovation on models and on training. But do not forget about the Algorithm Speed Prize! Speed is very valuable as well. Please feel free to ask us any questions regarding the data, the competition, our R&D team or our company.
\n\nWe hope you will have a lot of fun with this competition!\nLooking forward to hearing from you in the discussion!
\n\nThe R&D team from Airbus DS Intelligence.
", "rawMarkdown": "Dear Kagglers,\n\nOn behalf of Airbus DS Intelligence and our R&D team, I would like to officially welcome you to the “Airbus Ship Detection Challenge”!\n\nThis competition is about finding ships on satellite images as quickly as possible. To fight against potentially devastating events at sea like ship accidents, piracy or illegal fishing, we need to act fast. This is why we challenge you to help us detect ALL ships on our satellite images with machine learning and do it FAST enough that we can send the alert on time.\n\nWe have provided a large quantity of labeled images from our constellation of SPOT satellites as well as an even bigger collection of images of sea without ships. This is because finding ships in high seas is like finding a needle in a haystack! I am sure that this will generate a lot of innovation on models and on training. But do not forget about the Algorithm Speed Prize! Speed is very valuable as well. Please feel free to ask us any questions regarding the data, the competition, our R&D team or our company. \n\nWe hope you will have a lot of fun with this competition!\nLooking forward to hearing from you in the discussion! \n\nThe R&D team from Airbus DS Intelligence.", "votes": 21 }, { "id": 368377, "postDate": "2018-08-09T19:36:39.507Z", "content": "This seems to be very interesting, but I am wondering if I am helping out the military here? Airbus is one of the biggest defence companies in Europe. Identify vessels quickly on the sea can be very helpful.
", "rawMarkdown": "This seems to be very interesting, but I am wondering if I am helping out the military here? Airbus is one of the biggest defence companies in Europe. Identify vessels quickly on the sea can be very helpful. ", "votes": 10 }, { "id": 364322, "postDate": "2018-07-31T10:05:05.060Z", "content": "Hi Jeff, first of all thank you for organising this competition! It will be interesting to learn more about the current solutions used at Airbus and more specifically the scores the R&D team is able to achieve on this dataset or a similar dataset in production. I think releasing such a benchmark at the beginning of the competition can be interesting, because it gives an idea of what is already done and possible and might challenge us Kagglers even more to push harder.
", "rawMarkdown": "Hi Jeff, first of all thank you for organising this competition! It will be interesting to learn more about the current solutions used at Airbus and more specifically the scores the R&D team is able to achieve on this dataset or a similar dataset in production. I think releasing such a benchmark at the beginning of the competition can be interesting, because it gives an idea of what is already done and possible and might challenge us Kagglers even more to push harder.", "votes": 7, "replies": [ { "id": 364372, "postDate": "2018-07-31T11:50:21.583Z", "content": "Hi Indra, thank you for your comment and suggestion. We will look into that. Cheers, Jeff.
", "rawMarkdown": "Hi Indra, thank you for your comment and suggestion. We will look into that. Cheers, Jeff.", "votes": 2 } ] }, { "id": 366289, "postDate": "2018-08-04T17:03:24.200Z", "content": "Hello Jeff,
\n\nRegarding the Algorithm Speed contest, the size of the 40 000 chips will be as well 768x768 pixels ?
\n\nAlexis
", "rawMarkdown": "Hello Jeff,\n\nRegarding the Algorithm Speed contest, the size of the 40 000 chips will be as well 768x768 pixels ?\n\nAlexis", "votes": 1, "replies": [ { "id": 368119, "postDate": "2018-08-09T09:45:42.213Z", "content": "Hi Alexis,
\n\nYes, it will be the same size.
\n\nKind regards,
\n\nJeff.
", "rawMarkdown": "Hi Alexis,\n\nYes, it will be the same size. \n\nKind regards,\n\nJeff.", "votes": 1 } ] }, { "id": 2472192, "postDate": "2023-10-07T05:31:39.180Z", "content": "Mr. Faudi, would I be able to use the data set in a graduate research project? I am working on creating a deep convolutional neural network for my Computer Vision course. Thank you.
", "rawMarkdown": "Mr. Faudi, would I be able to use the data set in a graduate research project? I am working on creating a deep convolutional neural network for my Computer Vision course. Thank you." }, { "id": 399891, "postDate": "2018-10-07T03:28:30.957Z", "content": "how the inference speed gets tested? Are you using a GPU or CPU for speed test?
\n\nEspecially, if I can optimize everything to run on GPU (and I would, therefore, use the NCHW - Format) or if we are bound to use CPU-Kernels for speed?
", "rawMarkdown": "how the inference speed gets tested? Are you using a GPU or CPU for speed test? \n\nEspecially, if I can optimize everything to run on GPU (and I would, therefore, use the NCHW - Format) or if we are bound to use CPU-Kernels for speed?" }, { "id": 365649, "postDate": "2018-08-03T05:19:52.700Z", "content": "What do you mean by encoded pixels ?
", "rawMarkdown": "What do you mean by encoded pixels ?", "replies": [ { "id": 372871, "postDate": "2018-08-20T13:59:52.160Z", "content": "That means that binary mask that you have on the output should be encoded using RLE (Run Length Encoding). \nHere you can find how Run Length Encoding and Decoding are implemented:\nhttps://www.kaggle.com/paulorzp/run-length-encode-and-decode
", "rawMarkdown": "That means that binary mask that you have on the output should be encoded using RLE (Run Length Encoding). \nHere you can find how Run Length Encoding and Decoding are implemented:\nhttps://www.kaggle.com/paulorzp/run-length-encode-and-decode" } ] }, { "id": 364735, "postDate": "2018-08-01T06:17:44.913Z", "content": "Looks really challenging hopefully it will provide an opportunity to leanrn from veterans here in Kaggle.
", "rawMarkdown": "Looks really challenging hopefully it will provide an opportunity to leanrn from veterans here in Kaggle." }, { "id": 364259, "postDate": "2018-07-31T06:35:59.887Z", "rawMarkdown": "", "isDeleted": true }, { "id": 364208, "postDate": "2018-07-31T02:03:35.420Z", "rawMarkdown": "", "isDeleted": true } ], "comments": [ { "id": 368377, "author_name": "ClemensF", "author_url": "", "post_date": "2018-08-09T19:36:39.507000", "content": "This seems to be very interesting, but I am wondering if I am helping out the military here? Airbus is one of the biggest defence companies in Europe. Identify vessels quickly on the sea can be very helpful.
", "votes": 10, "replies": [] }, { "id": 364322, "author_name": "Indra den Bakker", "author_url": "", "post_date": "2018-07-31T10:05:05.060000", "content": "Hi Jeff, first of all thank you for organising this competition! It will be interesting to learn more about the current solutions used at Airbus and more specifically the scores the R&D team is able to achieve on this dataset or a similar dataset in production. I think releasing such a benchmark at the beginning of the competition can be interesting, because it gives an idea of what is already done and possible and might challenge us Kagglers even more to push harder.
", "votes": 7, "replies": [ { "id": 364372, "author_name": "Jeff Faudi", "author_url": "", "post_date": "2018-07-31T11:50:21.583000", "content": "Hi Indra, thank you for your comment and suggestion. We will look into that. Cheers, Jeff.
", "votes": 2, "replies": [] } ] }, { "id": 366289, "author_name": "Alexis Letulier", "author_url": "", "post_date": "2018-08-04T17:03:24.200000", "content": "Hello Jeff,
\n\nRegarding the Algorithm Speed contest, the size of the 40 000 chips will be as well 768x768 pixels ?
\n\nAlexis
", "votes": 1, "replies": [ { "id": 368119, "author_name": "Jeff Faudi", "author_url": "", "post_date": "2018-08-09T09:45:42.213000", "content": "Hi Alexis,
\n\nYes, it will be the same size.
\n\nKind regards,
\n\nJeff.
", "votes": 1, "replies": [] } ] }, { "id": 2472192, "author_name": "Bill Gonzalez", "author_url": "", "post_date": "2023-10-07T05:31:39.180000", "content": "Mr. Faudi, would I be able to use the data set in a graduate research project? I am working on creating a deep convolutional neural network for my Computer Vision course. Thank you.
", "votes": 0, "replies": [] }, { "id": 399891, "author_name": "Manuel Cueva", "author_url": "", "post_date": "2018-10-07T03:28:30.957000", "content": "how the inference speed gets tested? Are you using a GPU or CPU for speed test?
\n\nEspecially, if I can optimize everything to run on GPU (and I would, therefore, use the NCHW - Format) or if we are bound to use CPU-Kernels for speed?
", "votes": 0, "replies": [] }, { "id": 365649, "author_name": "SSJohn", "author_url": "", "post_date": "2018-08-03T05:19:52.700000", "content": "What do you mean by encoded pixels ?
", "votes": 0, "replies": [ { "id": 372871, "author_name": "Orest Rehusevych", "author_url": "", "post_date": "2018-08-20T13:59:52.160000", "content": "That means that binary mask that you have on the output should be encoded using RLE (Run Length Encoding). \nHere you can find how Run Length Encoding and Decoding are implemented:\nhttps://www.kaggle.com/paulorzp/run-length-encode-and-decode
", "votes": 0, "replies": [] } ] }, { "id": 364735, "author_name": "Deekshith", "author_url": "", "post_date": "2018-08-01T06:17:44.913000", "content": "Looks really challenging hopefully it will provide an opportunity to leanrn from veterans here in Kaggle.
", "votes": 0, "replies": [] }, { "id": 364259, "author_name": "", "author_url": "", "post_date": "2018-07-31T06:35:59.887000", "content": "", "votes": 0, "replies": [] }, { "id": 364208, "author_name": "", "author_url": "", "post_date": "2018-07-31T02:03:35.420000", "content": "", "votes": 0, "replies": [] } ], "raw_markdown_by_id": { "364140": "Dear Kagglers,\n\nOn behalf of Airbus DS Intelligence and our R&D team, I would like to officially welcome you to the “Airbus Ship Detection Challenge”!\n\nThis competition is about finding ships on satellite images as quickly as possible. To fight against potentially devastating events at sea like ship accidents, piracy or illegal fishing, we need to act fast. This is why we challenge you to help us detect ALL ships on our satellite images with machine learning and do it FAST enough that we can send the alert on time.\n\nWe have provided a large quantity of labeled images from our constellation of SPOT satellites as well as an even bigger collection of images of sea without ships. This is because finding ships in high seas is like finding a needle in a haystack! I am sure that this will generate a lot of innovation on models and on training. But do not forget about the Algorithm Speed Prize! Speed is very valuable as well. Please feel free to ask us any questions regarding the data, the competition, our R&D team or our company. \n\nWe hope you will have a lot of fun with this competition!\nLooking forward to hearing from you in the discussion! \n\nThe R&D team from Airbus DS Intelligence.", "368377": "This seems to be very interesting, but I am wondering if I am helping out the military here? Airbus is one of the biggest defence companies in Europe. Identify vessels quickly on the sea can be very helpful. ", "364322": "Hi Jeff, first of all thank you for organising this competition! It will be interesting to learn more about the current solutions used at Airbus and more specifically the scores the R&D team is able to achieve on this dataset or a similar dataset in production. I think releasing such a benchmark at the beginning of the competition can be interesting, because it gives an idea of what is already done and possible and might challenge us Kagglers even more to push harder.", "366289": "Hello Jeff,\n\nRegarding the Algorithm Speed contest, the size of the 40 000 chips will be as well 768x768 pixels ?\n\nAlexis", "2472192": "Mr. Faudi, would I be able to use the data set in a graduate research project? I am working on creating a deep convolutional neural network for my Computer Vision course. Thank you.", "399891": "how the inference speed gets tested? Are you using a GPU or CPU for speed test? \n\nEspecially, if I can optimize everything to run on GPU (and I would, therefore, use the NCHW - Format) or if we are bound to use CPU-Kernels for speed?", "365649": "What do you mean by encoded pixels ?", "364735": "Looks really challenging hopefully it will provide an opportunity to leanrn from veterans here in Kaggle.", "364259": "", "364208": "" } }