{ "id": 569704, "title": "Does satellite data on seafile have the same error", "url": "/competitions/geolifeclef-2025/discussion/569704", "author_name": "", "post_date": "2025-03-23T15:56:22.238579700Z", "votes": 1, "comment_count": 4, "views": 0, "content": "

Hi, I noticed that the seafile also has data from the test set on it. Can you tell me if the PO and PA_test datasets on the seafile have the same error? If there is an error, will the data on the seafile be updated during this challenge?

", "messages": [ { "id": "3157586", "postDate": "03/23/2025 15:56:22", "content": "

Hi, I noticed that the seafile also has data from the test set on it. Can you tell me if the PO and PA_test datasets on the seafile have the same error? If there is an error, will the data on the seafile be updated during this challenge?

", "rawMarkdown": "Hi, I noticed that the seafile also has data from the test set on it. Can you tell me if the PO and PA_test datasets on the seafile have the same error? If there is an error, will the data on the seafile be updated during this challenge?", "votes": null }, { "id": "3157638", "postDate": "03/23/2025 16:29:19", "content": "

Dear @liuhaixu,

\n

Yes, all the Sentinel2 data, no matter where you find it, have the problem.
\nWe are still working on the fix. Currently, 90% of the PA data is done. PO data will probably come later next week.
\nEven though you can use the old data, we do not recommend that since, for example, different normalization is used for training / test data processing.

\n

Best,
\nLukas

", "rawMarkdown": "Dear @liuhaixu,\n\nYes, all the Sentinel2 data, no matter where you find it, have the problem. \nWe are still working on the fix. Currently, 90% of the PA data is done. PO data will probably come later next week.\nEven though you can use the old data, we do not recommend that since, for example, different normalization is used for training / test data processing.\n\nBest,\nLukas", "votes": null }, { "id": "3158509", "postDate": "03/24/2025 16:00:29", "content": "

Thank for you reply!

", "rawMarkdown": "Thank for you reply!", "votes": null }, { "id": "3159729", "postDate": "03/25/2025 21:25:28", "content": "

Dear @liuhaixu,

\n

The Sentinel-2 satellite patches for Presence-Absence (PA) are now available on Kaggle and Seafile.
\nGiven the problem I have mentioned earlier, we decided to switch from JPEG to TIFF.
\nSince it might be new for some participants, I have also created a Kaggle Notebook where I describe, process and visualise the data.

\n

Have Fun!

\n

Best,
\nLukas

", "rawMarkdown": "Dear @liuhaixu,\n\nThe Sentinel-2 satellite patches for Presence-Absence (PA) are now available on Kaggle and Seafile.\nGiven the problem I have mentioned earlier, we decided to switch from JPEG to TIFF. \nSince it might be new for some participants, I have also created a [Kaggle Notebook](https://www.kaggle.com/code/picekl/sentinel-2-data-processing-and-normalization) where I describe, process and visualise the data. \n\nHave Fun!\n\nBest,\nLukas", "votes": null }, { "id": "3159890", "postDate": "03/26/2025 03:34:49", "content": "

Good news! Thank you!

", "rawMarkdown": "Good news! Thank you!", "votes": null } ], "comments": [ { "id": 3157638, "author_name": "picekl", "author_url": "", "post_date": "03/23/2025 16:29:19", "content": "

Dear @liuhaixu,

\n

Yes, all the Sentinel2 data, no matter where you find it, have the problem.
\nWe are still working on the fix. Currently, 90% of the PA data is done. PO data will probably come later next week.
\nEven though you can use the old data, we do not recommend that since, for example, different normalization is used for training / test data processing.

\n

Best,
\nLukas

", "votes": null, "replies": [ { "id": 3158509, "author_name": "liuhaixu", "author_url": "", "post_date": "03/24/2025 16:00:29", "content": "

Thank for you reply!

", "votes": null, "replies": [ { "id": 3159729, "author_name": "picekl", "author_url": "", "post_date": "03/25/2025 21:25:28", "content": "

Dear @liuhaixu,

\n

The Sentinel-2 satellite patches for Presence-Absence (PA) are now available on Kaggle and Seafile.
\nGiven the problem I have mentioned earlier, we decided to switch from JPEG to TIFF.
\nSince it might be new for some participants, I have also created a Kaggle Notebook where I describe, process and visualise the data.

\n

Have Fun!

\n

Best,
\nLukas

", "votes": null, "replies": [ { "id": 3159890, "author_name": "liuhaixu", "author_url": "", "post_date": "03/26/2025 03:34:49", "content": "

Good news! Thank you!

", "votes": null, "replies": [] } ] } ] } ] } ], "raw_markdown_by_id": { "3157586": "Hi, I noticed that the seafile also has data from the test set on it. Can you tell me if the PO and PA_test datasets on the seafile have the same error? If there is an error, will the data on the seafile be updated during this challenge?", "3157638": "Dear @liuhaixu,\n\nYes, all the Sentinel2 data, no matter where you find it, have the problem. \nWe are still working on the fix. Currently, 90% of the PA data is done. PO data will probably come later next week.\nEven though you can use the old data, we do not recommend that since, for example, different normalization is used for training / test data processing.\n\nBest,\nLukas", "3158509": "Thank for you reply!", "3159729": "Dear @liuhaixu,\n\nThe Sentinel-2 satellite patches for Presence-Absence (PA) are now available on Kaggle and Seafile.\nGiven the problem I have mentioned earlier, we decided to switch from JPEG to TIFF. \nSince it might be new for some participants, I have also created a [Kaggle Notebook](https://www.kaggle.com/code/picekl/sentinel-2-data-processing-and-normalization) where I describe, process and visualise the data. \n\nHave Fun!\n\nBest,\nLukas", "3159890": "Good news! Thank you!" }, "source": "meta" }