Robust ecological analysis of camera trap data labelled by a machine learning model
Problem
Labelling the millions of images generated by camera trap arrays is too slow, causing long delays between data collection and ecological inference, which impedes conservation.
Why it matters
Delayed species data slows conservation decisions during what the authors call a time of ecological crisis.
What the AI did
The authors built, tested and applied a machine learning model to classify 26 Central African forest mammal and bird species (or groups) from camera trap images, aiming at fully automated ecological analyses.
Results
Model built to classify 26 species/groups; performance metrics not stated in source (abstract truncated)
Impact
The approach aims to remove human labelling latency from camera trap ecological analyses.
Record
| Venue | Methods in Ecology and Evolution |
|---|---|
| Published | 2021 |
| Work type | article |
| Authors | Robin C. Whytock, Jędrzej Świeżewski, Joeri A. Zwerts, Tadeusz Bara‐Słupski, Aurélie Flore Koumba Pambo, Marek Rogala, Laila Bahaa‐el‐din, Kelly Boekee et al. |
| Citations | 89 (OpenAlex, at import time) |
| References | 40 |
| Model or method | not stated in source |
| Organization | not stated in source |
| Primary source | https://doi.org/10.1111/2041-210x.13576 |
| Open access | hybrid |
| Keywords | camera traps, machine learning, species classification, conservation, automated labelling |
Importance rating
significant — thousands to millions affected
Abstract states camera trap arrays generate millions of images and labelling latency impedes conservation at a time of ecological crisis; broad relevance to ecological monitoring.