Welcome to the AnimalCLEF25! 🐢🦎🐆
\nAnimalCLEF is a new addition to the LifeCLEF lab that focuses on the evaluation of machine learning methods for not just species recognition but also for identifying individual animals. Therefore, your task is to design a model that, for each image, determines whether the depicted animal is new (not present in the training dataset) or known (in which case, its identity must be provided).
\nThis challenge simulates real-world conservation and biodiversity monitoring scenarios, where identifying previously unseen species is a key aspect of wildlife research. It requires the development of models that can handle the complexity of novelty detection and few-shot learning, critical skills for understanding and monitoring biodiversity in changing environments.
\n👨🔬 Lukáš Picek @picekl is a postdoctoral researcher at INRIA [France] and a PI at the University of West Bohemia in Pilsen [Czechia], specializing in applied machine learning, including species recognition, species distribution modeling, and animal identification. Since 2021, he has co-organized the CVPR-FGVC and LifeCLEF workshops and is the lead organizer of the newly accepted CVPR tutorial on Animal Re-identification. As a competition chair at FGVC workshops, he has helped organize or independently led multiple competitions with thousands of participants on Kaggle.
\n🧑💻 Lukáš Adam @wildlifedatasets fedatasets is a senior scientist at the University of West Bohemia in Pilsen, Czech Republic. Lukáš works on animal re-identification with an emphasis on applying machine learning and mathematical models to sea turtle research. His second expertise lies in performing optimization, identification and control of power electronics devices.
\n🔬 Kostas Papafitsoros @KPapafitsoros - is a lecturer in Mathematical Data Science at Queen Mary University of London, with dual expertise in two distinct domains: Mathematical Imaging & Inverse Problems and Wildlife Conservation with a particular focus on sea turtle He is working on a wide range of topics from deep learning in imaging to sea turtle ecological studies while he is also considered an expert in sea turtle re-identification with a 18-year experience on the topic.
\n🦾 Vojtěch Čermák @vojtacermak is a computer vision researcher and PhD student at the Czech Technical University in Prague. With industry experience as a data scientist, he is passionate about bridging cutting-edge research with real-world applications. His current work focuses on leveraging recent advances in machine learning for the individual identification of animals across multiple species.
\nAs your hosts, we're here to help, so do not mind asking any questions! 😉
\nHave a great Experience! 🚀🐢 🦎 🐆
", "messages": [ { "id": "3154912", "postDate": "03/20/2025 14:11:31", "content": "Welcome to the AnimalCLEF25! 🐢🦎🐆
\nAnimalCLEF is a new addition to the LifeCLEF lab that focuses on the evaluation of machine learning methods for not just species recognition but also for identifying individual animals. Therefore, your task is to design a model that, for each image, determines whether the depicted animal is new (not present in the training dataset) or known (in which case, its identity must be provided).
\nThis challenge simulates real-world conservation and biodiversity monitoring scenarios, where identifying previously unseen species is a key aspect of wildlife research. It requires the development of models that can handle the complexity of novelty detection and few-shot learning, critical skills for understanding and monitoring biodiversity in changing environments.
\n👨🔬 Lukáš Picek @picekl is a postdoctoral researcher at INRIA [France] and a PI at the University of West Bohemia in Pilsen [Czechia], specializing in applied machine learning, including species recognition, species distribution modeling, and animal identification. Since 2021, he has co-organized the CVPR-FGVC and LifeCLEF workshops and is the lead organizer of the newly accepted CVPR tutorial on Animal Re-identification. As a competition chair at FGVC workshops, he has helped organize or independently led multiple competitions with thousands of participants on Kaggle.
\n🧑💻 Lukáš Adam @wildlifedatasets fedatasets is a senior scientist at the University of West Bohemia in Pilsen, Czech Republic. Lukáš works on animal re-identification with an emphasis on applying machine learning and mathematical models to sea turtle research. His second expertise lies in performing optimization, identification and control of power electronics devices.
\n🔬 Kostas Papafitsoros @KPapafitsoros - is a lecturer in Mathematical Data Science at Queen Mary University of London, with dual expertise in two distinct domains: Mathematical Imaging & Inverse Problems and Wildlife Conservation with a particular focus on sea turtle He is working on a wide range of topics from deep learning in imaging to sea turtle ecological studies while he is also considered an expert in sea turtle re-identification with a 18-year experience on the topic.
\n🦾 Vojtěch Čermák @vojtacermak is a computer vision researcher and PhD student at the Czech Technical University in Prague. With industry experience as a data scientist, he is passionate about bridging cutting-edge research with real-world applications. His current work focuses on leveraging recent advances in machine learning for the individual identification of animals across multiple species.
\nAs your hosts, we're here to help, so do not mind asking any questions! 😉
\nHave a great Experience! 🚀🐢 🦎 🐆
", "rawMarkdown": "**Welcome to the AnimalCLEF25!** 🐢🦎🐆\n\nAnimalCLEF is a new addition to the **LifeCLEF lab** that focuses on the evaluation of machine learning methods for not just species recognition but also for identifying individual animals. Therefore, your task is to design a model that, for each image, determines whether the depicted animal is **new** (not present in the training dataset) or **known** (in which case, its identity must be provided).\n\n### Why is this important?\nThis challenge simulates real-world conservation and biodiversity monitoring scenarios, where identifying previously unseen species is a key aspect of wildlife research. It requires the development of models that can handle the complexity of novelty detection and few-shot learning, critical skills for understanding and monitoring biodiversity in changing environments.\n\n## How are we?\n\n👨🔬 **Lukáš Picek** @picekl is a postdoctoral researcher at INRIA [_France_] and a PI at the University of West Bohemia in Pilsen [_Czechia_], specializing in applied machine learning, including species recognition, species distribution modeling, and animal identification. Since 2021, he has co-organized the CVPR-FGVC and LifeCLEF workshops and is the lead organizer of the newly accepted CVPR tutorial on Animal Re-identification. As a competition chair at FGVC workshops, he has helped organize or independently led multiple competitions with thousands of participants on Kaggle.\n\n🧑💻 **Lukáš Adam** @wildlifedatasets fedatasets is a senior scientist at the University of West Bohemia in Pilsen, Czech Republic. Lukáš works on animal re-identification with an emphasis on applying machine learning and mathematical models to sea turtle research. His second expertise lies in performing optimization, identification and control of power electronics devices.\n\n🔬 **Kostas Papafitsoros** [@KPapafitsoros](https://www.kaggle.com/KPapafitsoros) - is a lecturer in Mathematical Data Science at Queen Mary University of London, with dual expertise in two distinct domains: Mathematical Imaging & Inverse Problems and Wildlife Conservation with a particular focus on sea turtle He is working on a wide range of topics from deep learning in imaging to sea turtle ecological studies while he is also considered an expert in sea turtle re-identification with a 18-year experience on the topic.\n\n🦾 **Vojtěch Čermák** @vojtacermak is a computer vision researcher and PhD student at the Czech Technical University in Prague. With industry experience as a data scientist, he is passionate about bridging cutting-edge research with real-world applications. His current work focuses on leveraging recent advances in machine learning for the individual identification of animals across multiple species.\n\nAs your hosts, we're here to help, so do not mind asking any questions! 😉 \n\n**Have a great Experience!** 🚀🐢 🦎 🐆", "votes": null }, { "id": "3156343", "postDate": "03/22/2025 03:48:52", "content": "Thanks for hosting the competition. I would like to know what's the criteria for writing the paper? Should our team be 1st in leaderboard to do so?
", "rawMarkdown": "Thanks for hosting the competition. I would like to know what's the criteria for writing the paper? Should our team be 1st in leaderboard to do so?", "votes": null }, { "id": "3157338", "postDate": "03/23/2025 09:55:22", "content": "Hi @vishakkbhat,
\nAs in previous years, we encourage all participants to contribute by publishing working note papers detailing their solutions even if not among the top-scoring participants!
\nIf you have an interesting method, ablation study, or novel insight—we want to hear from you!
\nPS: I tried to put more detail into this post.
\nBest,
\nLukas
Thanks @picekl, that answers my question.
", "rawMarkdown": "Thanks @picekl, that answers my question.", "votes": null } ], "comments": [ { "id": 3156343, "author_name": "vishakkbhat", "author_url": "", "post_date": "03/22/2025 03:48:52", "content": "Thanks for hosting the competition. I would like to know what's the criteria for writing the paper? Should our team be 1st in leaderboard to do so?
", "votes": null, "replies": [ { "id": 3157338, "author_name": "picekl", "author_url": "", "post_date": "03/23/2025 09:55:22", "content": "Hi @vishakkbhat,
\nAs in previous years, we encourage all participants to contribute by publishing working note papers detailing their solutions even if not among the top-scoring participants!
\nIf you have an interesting method, ablation study, or novel insight—we want to hear from you!
\nPS: I tried to put more detail into this post.
\nBest,
\nLukas
Thanks @picekl, that answers my question.
", "votes": null, "replies": [] } ] } ] } ], "raw_markdown_by_id": { "3154912": "**Welcome to the AnimalCLEF25!** 🐢🦎🐆\n\nAnimalCLEF is a new addition to the **LifeCLEF lab** that focuses on the evaluation of machine learning methods for not just species recognition but also for identifying individual animals. Therefore, your task is to design a model that, for each image, determines whether the depicted animal is **new** (not present in the training dataset) or **known** (in which case, its identity must be provided).\n\n### Why is this important?\nThis challenge simulates real-world conservation and biodiversity monitoring scenarios, where identifying previously unseen species is a key aspect of wildlife research. It requires the development of models that can handle the complexity of novelty detection and few-shot learning, critical skills for understanding and monitoring biodiversity in changing environments.\n\n## How are we?\n\n👨🔬 **Lukáš Picek** @picekl is a postdoctoral researcher at INRIA [_France_] and a PI at the University of West Bohemia in Pilsen [_Czechia_], specializing in applied machine learning, including species recognition, species distribution modeling, and animal identification. Since 2021, he has co-organized the CVPR-FGVC and LifeCLEF workshops and is the lead organizer of the newly accepted CVPR tutorial on Animal Re-identification. As a competition chair at FGVC workshops, he has helped organize or independently led multiple competitions with thousands of participants on Kaggle.\n\n🧑💻 **Lukáš Adam** @wildlifedatasets fedatasets is a senior scientist at the University of West Bohemia in Pilsen, Czech Republic. Lukáš works on animal re-identification with an emphasis on applying machine learning and mathematical models to sea turtle research. His second expertise lies in performing optimization, identification and control of power electronics devices.\n\n🔬 **Kostas Papafitsoros** [@KPapafitsoros](https://www.kaggle.com/KPapafitsoros) - is a lecturer in Mathematical Data Science at Queen Mary University of London, with dual expertise in two distinct domains: Mathematical Imaging & Inverse Problems and Wildlife Conservation with a particular focus on sea turtle He is working on a wide range of topics from deep learning in imaging to sea turtle ecological studies while he is also considered an expert in sea turtle re-identification with a 18-year experience on the topic.\n\n🦾 **Vojtěch Čermák** @vojtacermak is a computer vision researcher and PhD student at the Czech Technical University in Prague. With industry experience as a data scientist, he is passionate about bridging cutting-edge research with real-world applications. His current work focuses on leveraging recent advances in machine learning for the individual identification of animals across multiple species.\n\nAs your hosts, we're here to help, so do not mind asking any questions! 😉 \n\n**Have a great Experience!** 🚀🐢 🦎 🐆", "3156343": "Thanks for hosting the competition. I would like to know what's the criteria for writing the paper? Should our team be 1st in leaderboard to do so?", "3157338": "Hi @vishakkbhat,\n\nAs in previous years, we encourage all participants to contribute by publishing working note papers detailing their solutions **even if not among the top-scoring participants!** \n\nIf you have an interesting method, ablation study, or novel insight—**we want to hear from you!** \n\nPS: I tried to put more detail into [this post](https://www.kaggle.com/competitions/animal-clef-2025/discussion/569663).\n\nBest,\nLukas", "3157966": "Thanks @picekl, that answers my question." }, "source": "meta" }