Devis Tuia

This study evaluates multi-scale deep learning models for species distribution, showing improved performance and reduced sensitivity to scale selection across plant and animal species.

This study develops a time-aware deep learning model to predict seasonal shifts in the monarch butterfly's niche under climate change, finding a northwestward range shift and contraction in key overwintering sites.

WildCLIP retrieves camera trap images based on environmental and ecological attributes using a domain-adapted vision-language model.