Track genomics, neuroscience and molecular biology.
Search papers, labs, and topics across Lattice.
The authors evaluated Vision Transformers (DINOv2, BioCLIP 2) against modern CNNs (ConvNeXt V2-L, EfficientNetV2-L) on a species-stratified benchmark of 16,701 photographs across 1,350 New Guinea orchid species to tackle extreme low-data, fine-grained identification. DINOv2 emerged as the top performer, achieving 66.9% macro top-1 genus accuracy and 86.6% species Recall@5, while outperforming domain-matched BioCLIP 2 by 7.1 percentage points and demonstrating strong open-set novel genus detection (0.958 AUROC). The findings demonstrate that generic self-supervised vision representations provide superior feature geometry over domain-specific contrastive models for fine-grained classification across long-tailed, data-starved natural distributions.
General-purpose self-supervised vision models outperform specialized, domain-matched biological foundation models by over 7 macro top-1 points when classifying hyperdiverse, data-starved visual long tails.
New Guinea is the world's richest island flora (~2,856 orchid species), yet most species are represented by only a handful of photographs, far fewer than direct species-level classification requires. Methods for fine-grained identification in such species-rich, data-poor floras are needed, and it remains unclear which backbone architecture and pretraining strategy best support them. We built a two-stage system that first predicts the genus of a query photograph, then retrieves visually similar reference images of candidate species using FAISS. We compared four pretrained backbones -- two Vision Transformers (ViTs; DINOv2, BioCLIP 2) and two CNNs (ConvNeXt V2-L, EfficientNetV2-L) -- fine-tuned under an identical protocol on a fixed, species-stratified partition of 16,701 photographs spanning 120 genera and 1,350 species, assessing accuracy, calibration, error structure, species retrieval, and open-set detection of novel genera. DINOv2 attained the best genus performance (macro top-1 66.9%, 95% CI 63.7-70.6; global top-1 88.9%); both ViTs outranked both CNNs, and general-purpose self-supervised pretraining (DINOv2) outperformed domain-matched biological pretraining (BioCLIP 2) by 7.1 points of macro top-1. Errors concentrated on two abundant genera acting as error attractors. DINOv2 embeddings achieved species Recall@5 of 86.6% and genus Recall@5 of 98.7%; temperature scaling reduced every backbone's Expected Calibration Error to about 0.03; and a distance-based open-set gate flagged unseen genera (mean AUROC 0.958). A self-supervised Vision-Transformer backbone combined with embedding retrieval is an effective, deployable strategy for fine-grained identification in species-rich, data-poor floras. The system is released as an open web application (the New Guinea Orchid Identifier), offering a practical template for other hyperdiverse, under-documented taxa.