Saved in:
Bibliographic Details
Main Authors: Leblanc, César, Picek, Lukas, Deneu, Benjamin, Bonnet, Pierre, Servajean, Maximilien, Palard, Rémi, Joly, Alexis
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2504.05231
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910905491521536
author Leblanc, César
Picek, Lukas
Deneu, Benjamin
Bonnet, Pierre
Servajean, Maximilien
Palard, Rémi
Joly, Alexis
author_facet Leblanc, César
Picek, Lukas
Deneu, Benjamin
Bonnet, Pierre
Servajean, Maximilien
Palard, Rémi
Joly, Alexis
contents This paper describes a cascading multimodal pipeline for high-resolution biodiversity mapping across Europe, integrating species distribution modeling, biodiversity indicators, and habitat classification. The proposed pipeline first predicts species compositions using a deep-SDM, a multimodal model trained on remote sensing, climate time series, and species occurrence data at 50x50m resolution. These predictions are then used to generate biodiversity indicator maps and classify habitats with Pl@ntBERT, a transformer-based LLM designed for species-to-habitat mapping. With this approach, continental-scale species distribution maps, biodiversity indicator maps, and habitat maps are produced, providing fine-grained ecological insights. Unlike traditional methods, this framework enables joint modeling of interspecies dependencies, bias-aware training with heterogeneous presence-absence data, and large-scale inference from multi-source remote sensing inputs.
format Preprint
id arxiv_https___arxiv_org_abs_2504_05231
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Mapping biodiversity at very-high resolution in Europe
Leblanc, César
Picek, Lukas
Deneu, Benjamin
Bonnet, Pierre
Servajean, Maximilien
Palard, Rémi
Joly, Alexis
Artificial Intelligence
Computer Vision and Pattern Recognition
Machine Learning
This paper describes a cascading multimodal pipeline for high-resolution biodiversity mapping across Europe, integrating species distribution modeling, biodiversity indicators, and habitat classification. The proposed pipeline first predicts species compositions using a deep-SDM, a multimodal model trained on remote sensing, climate time series, and species occurrence data at 50x50m resolution. These predictions are then used to generate biodiversity indicator maps and classify habitats with Pl@ntBERT, a transformer-based LLM designed for species-to-habitat mapping. With this approach, continental-scale species distribution maps, biodiversity indicator maps, and habitat maps are produced, providing fine-grained ecological insights. Unlike traditional methods, this framework enables joint modeling of interspecies dependencies, bias-aware training with heterogeneous presence-absence data, and large-scale inference from multi-source remote sensing inputs.
title Mapping biodiversity at very-high resolution in Europe
topic Artificial Intelligence
Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2504.05231