Tricho-Vision

Tricho-Vision applies computer vision to trichotaxonomy — the classification of hair/trichome microstructures — to enhance wildlife conservation of priority species. Published in Ecological Informatics (2025).

Motivation

Manual identification of species via trichome analysis is time-consuming and requires expert knowledge. Automated vision pipelines can scale species identification from microscopy images, aiding conservation monitoring at scale.

Contributions

• Curated a novel dataset of trichome microscopy images across multiple priority species
• Benchmarked CNN and transformer-based classifiers on fine-grained trichome morphology
• Demonstrated that vision models match expert-level accuracy on held-out species

Publication

Das, A., Banerjee, P., Biswas, S. et al. Tricho-Vision: The use of computer vision in trichotaxonomy for enhancing wildlife conservation of priority species. Ecological Informatics, 2025.