Brandon Claridge @ichsofna.bsky.social · Mar 31 2. The coding involved with training, testing, and deploying a model is trivial. The real problem is capturing training images. That can be done very rapidly if you have access to enough authoritatively identified specimens.
2 likes 1 replies
?
Replies
3. Habitus photos will probably work great but perhaps closer shots will be needed to discriminate species/ taxa at lower taxonomic levels. Which shots are needed will likely differ taxon to taxon. #parasitoid #specimen #insectcollection