We’re back and more ambitious than ever! In this 2025 edition, GeoLifeCLEF is challenging you to push the boundaries of geolocation-based species identification. While we keep the same training data as in the previous year, we expand the test set to many new locations with no or limited Presence Absence (PA) surveys. Therefore, we expect you to further explore the usage of the available Presence Only (PO) data.
\nAccurate species identification from geo-tagged data is crucial for biodiversity research, conservation, and ecological studies. Your innovative approaches can help scientists and conservationists better monitor and protect life on Earth. This year’s dataset is packed with rich geospatial data, species observations, and environmental metadata, giving you the perfect playground to explore context-aware prediction and large-scale biodiversity modeling.
\nAs always, we’re here to support you—just not with the ground truth labels! 😉
\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🧑💻 Théo Larcher @tlarcher is a Ph.D. student at INRIA, Montpellier, who focuses on machine learning applied to biodiversity characterization, particularly in the context of multi-scale deep representations and species distribution modeling. He is the lead developper of deep-SDM framework 'Malpolon' which aims at facilitating the training of deep-learning PyTorch-based models for Species Distribution Modeling.
\n🌿 César Leblanc @leblanccesar is a Ph.D. student at INRIA, Montpellier. César works on developing advanced models for ecological data analysis, with a focus on improving species identification using geospatial and environmental data.
\n🌎 Maximilien Servajean @maximiliense is an Associate Professor at Université Paul Valéry Montpellier and researcher at LIRMM. Maximilien specializes in ecological modeling and biodiversity informatics, with a focus on species distribution modeling and ecological data analysis. His work leverages machine learning and statistical modeling to better understand biodiversity patterns and support conservation efforts. He’s also involved in open science initiatives, advocating for reproducible research and open ecological data. Maximilien contributes his expertise to the development of innovative methods for large-scale species prediction and geospatial data analysis.
\n🌱 Alexis Joly @alxjoly is a Research Director at INRIA, Montpellier. Alexis leads the PlantNet project, one of the world’s largest citizen science platforms for plant biodiversity monitoring, with over 20 million users worldwide. He’s the co-founder and scientific coordinator of LifeCLEF, an international evaluation campaign advancing AI-powered biodiversity identification research. His work bridges machine learning, ecological informatics, and citizen science to support biodiversity monitoring and conservation at a global scale.
\n🚀🌍 Have an amazing experience at GeoLifeCLEF 2025! 🌍🚀
", "messages": [ { "id": 3157340, "postDate": "2025-03-23T10:09:51.917Z", "content": "We’re back and more ambitious than ever! In this 2025 edition, GeoLifeCLEF is challenging you to push the boundaries of geolocation-based species identification. While we keep the same training data as in the previous year, we expand the test set to many new locations with no or limited Presence Absence (PA) surveys. Therefore, we expect you to further explore the usage of the available Presence Only (PO) data.
\nAccurate species identification from geo-tagged data is crucial for biodiversity research, conservation, and ecological studies. Your innovative approaches can help scientists and conservationists better monitor and protect life on Earth. This year’s dataset is packed with rich geospatial data, species observations, and environmental metadata, giving you the perfect playground to explore context-aware prediction and large-scale biodiversity modeling.
\nAs always, we’re here to support you—just not with the ground truth labels! 😉
\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🧑💻 Théo Larcher @tlarcher is a Ph.D. student at INRIA, Montpellier, who focuses on machine learning applied to biodiversity characterization, particularly in the context of multi-scale deep representations and species distribution modeling. He is the lead developper of deep-SDM framework 'Malpolon' which aims at facilitating the training of deep-learning PyTorch-based models for Species Distribution Modeling.
\n🌿 César Leblanc @leblanccesar is a Ph.D. student at INRIA, Montpellier. César works on developing advanced models for ecological data analysis, with a focus on improving species identification using geospatial and environmental data.
\n🌎 Maximilien Servajean @maximiliense is an Associate Professor at Université Paul Valéry Montpellier and researcher at LIRMM. Maximilien specializes in ecological modeling and biodiversity informatics, with a focus on species distribution modeling and ecological data analysis. His work leverages machine learning and statistical modeling to better understand biodiversity patterns and support conservation efforts. He’s also involved in open science initiatives, advocating for reproducible research and open ecological data. Maximilien contributes his expertise to the development of innovative methods for large-scale species prediction and geospatial data analysis.
\n🌱 Alexis Joly @alxjoly is a Research Director at INRIA, Montpellier. Alexis leads the PlantNet project, one of the world’s largest citizen science platforms for plant biodiversity monitoring, with over 20 million users worldwide. He’s the co-founder and scientific coordinator of LifeCLEF, an international evaluation campaign advancing AI-powered biodiversity identification research. His work bridges machine learning, ecological informatics, and citizen science to support biodiversity monitoring and conservation at a global scale.
\n🚀🌍 Have an amazing experience at GeoLifeCLEF 2025! 🌍🚀
", "rawMarkdown": "# 🌍🍃 Welcome to GeoLifeCLEF25! 🍃🌍\n\nWe’re back and more ambitious than ever! In this **2025 edition**, GeoLifeCLEF is challenging you to push the boundaries of **geolocation-based species identification**. While we keep the _same_ training data as in the previous year, **we expand the test set to many new locations with no or limited Presence Absence (PA) surveys**. Therefore, we expect you to further explore the usage of the available Presence Only (PO) data.\n\nAccurate species identification from geo-tagged data is crucial for biodiversity research, conservation, and ecological studies. Your innovative approaches can help scientists and conservationists **better monitor and protect life on Earth**. This year’s dataset is packed with **rich geospatial data, species observations, and environmental metadata**, giving you the perfect playground to explore **context-aware prediction and large-scale biodiversity modeling**.\n\nAs always, we’re here to support you—just not with the ground truth labels! 😉 \n\n## 🌟 Meet the Team 🌟\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🧑💻 **Théo Larcher** @tlarcher is a Ph.D. student at INRIA, Montpellier, who focuses on machine learning applied to biodiversity characterization, particularly in the context of multi-scale deep representations and species distribution modeling. He is the lead developper of deep-SDM framework 'Malpolon' which aims at facilitating the training of deep-learning PyTorch-based models for Species Distribution Modeling.\n\n🌿 **César Leblanc** @leblanccesar is a Ph.D. student at INRIA, Montpellier. César works on developing advanced models for ecological data analysis, with a focus on improving species identification using geospatial and environmental data.\n\n🌎 **Maximilien Servajean** @maximiliense is an Associate Professor at Université Paul Valéry Montpellier and researcher at LIRMM. Maximilien specializes in ecological modeling and biodiversity informatics, with a focus on species distribution modeling and ecological data analysis. His work leverages machine learning and statistical modeling to better understand biodiversity patterns and support conservation efforts. He’s also involved in open science initiatives, advocating for reproducible research and open ecological data. Maximilien contributes his expertise to the development of innovative methods for large-scale species prediction and geospatial data analysis.\n\n🌱 **Alexis Joly** @alxjoly is a Research Director at INRIA, Montpellier. Alexis leads the **PlantNet project**, one of the world’s largest citizen science platforms for plant biodiversity monitoring, with over 20 million users worldwide. He’s the co-founder and scientific coordinator of **LifeCLEF**, an international evaluation campaign advancing AI-powered biodiversity identification research. His work bridges machine learning, ecological informatics, and citizen science to support **biodiversity monitoring and conservation at a global scale**.\n\n\n🚀🌍 **Have an amazing experience at GeoLifeCLEF 2025!** 🌍🚀\n", "votes": 2 } ], "comments": [], "raw_markdown_by_id": { "3157340": "# 🌍🍃 Welcome to GeoLifeCLEF25! 🍃🌍\n\nWe’re back and more ambitious than ever! In this **2025 edition**, GeoLifeCLEF is challenging you to push the boundaries of **geolocation-based species identification**. While we keep the _same_ training data as in the previous year, **we expand the test set to many new locations with no or limited Presence Absence (PA) surveys**. Therefore, we expect you to further explore the usage of the available Presence Only (PO) data.\n\nAccurate species identification from geo-tagged data is crucial for biodiversity research, conservation, and ecological studies. Your innovative approaches can help scientists and conservationists **better monitor and protect life on Earth**. This year’s dataset is packed with **rich geospatial data, species observations, and environmental metadata**, giving you the perfect playground to explore **context-aware prediction and large-scale biodiversity modeling**.\n\nAs always, we’re here to support you—just not with the ground truth labels! 😉 \n\n## 🌟 Meet the Team 🌟\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🧑💻 **Théo Larcher** @tlarcher is a Ph.D. student at INRIA, Montpellier, who focuses on machine learning applied to biodiversity characterization, particularly in the context of multi-scale deep representations and species distribution modeling. He is the lead developper of deep-SDM framework 'Malpolon' which aims at facilitating the training of deep-learning PyTorch-based models for Species Distribution Modeling.\n\n🌿 **César Leblanc** @leblanccesar is a Ph.D. student at INRIA, Montpellier. César works on developing advanced models for ecological data analysis, with a focus on improving species identification using geospatial and environmental data.\n\n🌎 **Maximilien Servajean** @maximiliense is an Associate Professor at Université Paul Valéry Montpellier and researcher at LIRMM. Maximilien specializes in ecological modeling and biodiversity informatics, with a focus on species distribution modeling and ecological data analysis. His work leverages machine learning and statistical modeling to better understand biodiversity patterns and support conservation efforts. He’s also involved in open science initiatives, advocating for reproducible research and open ecological data. Maximilien contributes his expertise to the development of innovative methods for large-scale species prediction and geospatial data analysis.\n\n🌱 **Alexis Joly** @alxjoly is a Research Director at INRIA, Montpellier. Alexis leads the **PlantNet project**, one of the world’s largest citizen science platforms for plant biodiversity monitoring, with over 20 million users worldwide. He’s the co-founder and scientific coordinator of **LifeCLEF**, an international evaluation campaign advancing AI-powered biodiversity identification research. His work bridges machine learning, ecological informatics, and citizen science to support **biodiversity monitoring and conservation at a global scale**.\n\n\n🚀🌍 **Have an amazing experience at GeoLifeCLEF 2025!** 🌍🚀\n" } }