Tired of digitizing biodiversity data by hand? Consider using large language models to, for example, extract plant functional traits from unstructured texts: See bsapubs.onlinelibrary.wiley.com/doi/full/10.... led by @viktordo.bsky.social @uni-goettingen.de, @exeter.ac.uk, @ribesresearch.bsky.social

New dataset & pre-print: The Coralscapes dataset for semantic scene understanding in coral reefs! 🤿🎥🐟🐠🐡 Coralscapes is the first dataset consistently annotated for general-purpose semantic segmentation in reefs, spanning 174k segmentation masks of 39 benthic classes on 2074 images!