Automated facial recognition for wildlife that lack unique markings: A deep learning approach for brown bears
importance 4/10
solved
Wildlife Ecology and Conservation wildlife conservation stage: research open-source model
Problem
There was no objective visual method to identify individual animals of species lacking unique body markings, such as brown bears.
Why it matters
Individual identification underpins wildlife research and population monitoring that manual methods cannot scale.
What the AI did
A deep learning pipeline (object detection, landmark detection, similarity network, SVM classifier) detects bear faces and classifies individuals from images.
Results
not stated in source
Impact
The open-source BearID application enables automated individual identification for brown bears.
Record
| Venue | Ecology and Evolution |
|---|---|
| Published | 2020 |
| Work type | article |
| Authors | Melanie Clapham, Ed Miller, Mary Nguyen, Chris T. Darimont |
| Citations | 104 (OpenAlex, at import time) |
| References | 44 |
| Model or method | deep learning facial recognition (BearID) |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1002/ece3.6840 |
| Open access | gold |
| Keywords | facial recognition, deep learning, brown bears, wildlife monitoring, individual identification |
Importance rating
4
real but narrower advance
First objective visual ID method for species lacking markings, enabling wildlife research; niche but genuinely new capability, no scale stated.