The training data includes 4 types of \"State\" and 15 types of \"Species\".\nDoes the test data contain the same ones, or does it include any unknown types?
", "messages": [ { "id": 3312845, "postDate": "2025-11-08T05:29:47.877Z", "content": "The training data includes 4 types of \"State\" and 15 types of \"Species\".\nDoes the test data contain the same ones, or does it include any unknown types?
", "rawMarkdown": "The training data includes 4 types of \"State\" and 15 types of \"Species\".\nDoes the test data contain the same ones, or does it include any unknown types?", "votes": 21 }, { "id": 3312851, "postDate": "2025-11-08T05:45:04.840Z", "content": "Hi there, images in the testing set contain the same State and Species
", "rawMarkdown": "Hi there, images in the testing set contain the same State and Species", "votes": 15, "replies": [ { "id": 3312885, "postDate": "2025-11-08T06:45:35.567Z", "content": "Thank you for your prompt response! I understand now!
", "rawMarkdown": "Thank you for your prompt response! I understand now!" }, { "id": 3313308, "postDate": "2025-11-09T04:17:19.660Z", "content": "Does the test set contain the same sampling_date as train or does it include newer ones?
", "rawMarkdown": "Does the test set contain the same sampling_date as train or does it include newer ones?", "votes": 1, "replies": [ { "id": 3313419, "postDate": "2025-11-09T08:36:39.930Z", "content": "Not exactly the same, testing set should have contained some images from different sampling date.
", "rawMarkdown": "Not exactly the same, testing set should have contained some images from different sampling date.", "votes": 8 }, { "id": 3343400, "postDate": "2025-11-21T16:12:21.343Z", "content": "Since the paper describes data collection to have occurred starting in 2014 for 3 years - do we need to worry that the sampling came from an early or later year?
", "rawMarkdown": "Since the paper describes data collection to have occurred starting in 2014 for 3 years - do we need to worry that the sampling came from an early or later year?" }, { "id": 3355204, "postDate": "2025-11-30T14:59:28.757Z", "content": "@qiyuliaocsiro \nI understand that it was necessary to ensure that the test dataset does not contain images from the same sampling dates as in the training dataset in order to measure the generalization ability of the solutions across different times of the year. But I have a question:
\nwhen splitting the test dataset into public and private parts, was that completely random? Or did you apply split by sampling dates there too?
\nthanks.
", "rawMarkdown": "@qiyuliaocsiro \nI understand that it was necessary to ensure that the test dataset does not contain images from the same sampling dates as in the training dataset in order to measure the generalization ability of the solutions across different times of the year. But I have a question:\n\nwhen splitting the test dataset into public and private parts, was that completely random? Or did you apply split by sampling dates there too?\n\nthanks.", "votes": 2 }, { "id": 3360220, "postDate": "2025-12-02T01:38:24.423Z", "content": "It is not entirely random; the test set includes data collected at different time periods.
", "rawMarkdown": "It is not entirely random; the test set includes data collected at different time periods." }, { "id": 3377009, "postDate": "2025-12-15T14:19:39.807Z", "rawMarkdown": "", "isDeleted": true }, { "id": 3377272, "postDate": "2025-12-16T02:00:39.783Z", "rawMarkdown": "", "isDeleted": true }, { "id": 3378484, "postDate": "2025-12-18T10:23:50.947Z", "rawMarkdown": "", "isDeleted": true }, { "id": 3388226, "postDate": "2026-01-08T13:20:59.723Z", "rawMarkdown": "", "isDeleted": true } ] }, { "id": 3343253, "postDate": "2025-11-21T14:26:01.713Z", "content": "Just to clarify, am I correct in understanding that while the testing set has the same composition, we won't have access to that metadata during inference? Or is the metadata stored/available somewhere for us to use?\n(Translated by AI)
", "rawMarkdown": "Just to clarify, am I correct in understanding that while the testing set has the same composition, we won't have access to that metadata during inference? Or is the metadata stored/available somewhere for us to use?\n(Translated by AI)", "votes": -1, "replies": [ { "id": 3345931, "postDate": "2025-11-23T22:51:24.280Z", "content": "Metadata is not available in the testing phase
", "rawMarkdown": "Metadata is not available in the testing phase", "votes": 1 } ] } ] } ], "comments": [ { "id": 3312851, "author_name": "Qiyu Liao", "author_url": "", "post_date": "2025-11-08T05:45:04.840000", "content": "Hi there, images in the testing set contain the same State and Species
", "votes": 15, "replies": [ { "id": 3312885, "author_name": "hiro", "author_url": "", "post_date": "2025-11-08T06:45:35.567000", "content": "Thank you for your prompt response! I understand now!
", "votes": 0, "replies": [] }, { "id": 3313308, "author_name": "Med Ali Bouchhioua", "author_url": "", "post_date": "2025-11-09T04:17:19.660000", "content": "Does the test set contain the same sampling_date as train or does it include newer ones?
", "votes": 1, "replies": [ { "id": 3313419, "author_name": "Qiyu Liao", "author_url": "", "post_date": "2025-11-09T08:36:39.930000", "content": "Not exactly the same, testing set should have contained some images from different sampling date.
", "votes": 8, "replies": [] }, { "id": 3343400, "author_name": "PC Jimmmy", "author_url": "", "post_date": "2025-11-21T16:12:21.343000", "content": "Since the paper describes data collection to have occurred starting in 2014 for 3 years - do we need to worry that the sampling came from an early or later year?
", "votes": 0, "replies": [] }, { "id": 3355204, "author_name": "MD Mushfirat Mohaimin", "author_url": "", "post_date": "2025-11-30T14:59:28.757000", "content": "@qiyuliaocsiro \nI understand that it was necessary to ensure that the test dataset does not contain images from the same sampling dates as in the training dataset in order to measure the generalization ability of the solutions across different times of the year. But I have a question:
\nwhen splitting the test dataset into public and private parts, was that completely random? Or did you apply split by sampling dates there too?
\nthanks.
", "votes": 2, "replies": [] }, { "id": 3360220, "author_name": "Qiyu Liao", "author_url": "", "post_date": "2025-12-02T01:38:24.423000", "content": "It is not entirely random; the test set includes data collected at different time periods.
", "votes": 0, "replies": [] }, { "id": 3377009, "author_name": "", "author_url": "", "post_date": "2025-12-15T14:19:39.807000", "content": "", "votes": 0, "replies": [] }, { "id": 3377272, "author_name": "", "author_url": "", "post_date": "2025-12-16T02:00:39.783000", "content": "", "votes": 0, "replies": [] }, { "id": 3378484, "author_name": "", "author_url": "", "post_date": "2025-12-18T10:23:50.947000", "content": "", "votes": 0, "replies": [] }, { "id": 3388226, "author_name": "", "author_url": "", "post_date": "2026-01-08T13:20:59.723000", "content": "", "votes": 0, "replies": [] } ] }, { "id": 3343253, "author_name": "Atsushi_Matsumura", "author_url": "", "post_date": "2025-11-21T14:26:01.713000", "content": "Just to clarify, am I correct in understanding that while the testing set has the same composition, we won't have access to that metadata during inference? Or is the metadata stored/available somewhere for us to use?\n(Translated by AI)
", "votes": -1, "replies": [ { "id": 3345931, "author_name": "Qiyu Liao", "author_url": "", "post_date": "2025-11-23T22:51:24.280000", "content": "Metadata is not available in the testing phase
", "votes": 1, "replies": [] } ] } ] } ], "raw_markdown_by_id": { "3312845": "The training data includes 4 types of \"State\" and 15 types of \"Species\".\nDoes the test data contain the same ones, or does it include any unknown types?", "3312851": "Hi there, images in the testing set contain the same State and Species" } }