S M Rayeed
Home / Research

Animal re-identification

Population census depends on knowing which individual is which. I curated the Grevy's Zebra Census Dataset and helped develop multiview spatio-temporal track clustering for animal re-identification.

Camera views over timeview 1view 2view 3TrackletstimeSpatio-temporal clusteringzebra Azebra Bzebra CTracks that cannot co-occur in time are kept apart; the restare merged by appearance across views.
CVPR 2025 Workshop

Animal Re-Identification via Multiview Spatio-Temporal Track Clustering

S M Rayeed (co-author)
CVPR 2025 Workshop on CV4Animals

Individual animals are re-identified by clustering tracks across camera views and time instead of matching single frames, using spatio-temporal constraints to keep identities consistent.

Paper
Census photographs, one identity per individualGZ-014GZ-014GZ-027GZ-027GZ-031GZ-031GZ-042GZ-042Stripe patterns are unique to each zebra, so matching them across sightings gives a population count.
Dataset

GZCD: Grevy's Zebra Census Dataset

Hugging Face, Imageomics, CC BY 4.0

Census imagery of endangered Grevy's zebras for individual re-identification, detection and classification, with initial and final identity metadata. Grevy's zebras are endangered, and stripe-based identification lets a census count individuals without tagging them.

Open on Hugging Face

Other research areas

Fine-grained taxonomic identificationBiological visual question answeringKnowledge-based VQA and retrievalAI-ready biodiversity dataSign language recognition