A training protocol for human classification of Asian elephant images from trail cameras
- 1University of California San Diego
Abstract
Trail cameras have become ubiquitous tools for ecological data collection over recent decades. Despite progress in the development of automated algorithms and artificial intelligence for image classification, our ability to process large volumes of data remain limited by the need for trained human observers to make refined judgements. We provide guidance on placement of trail cameras for observing Asian elephants (Elephas maximus) and outline a protocol for training and testing naive human observers in performing image classifications (age/sex class and group composition) that cannot yet be automated. This process can be used to develop a high-throughput workflow capable of extracting useful data from large volumes of images. Our training material consisted of 14,007 images collected from 6 trail cameras around Udawalawe National Park in Sri Lanka from 2017-2019. In the first stage, expert observers (n=3) trained a group of inexperienced participants (n=4), who engaged in an iterative process to develop a protocol document. The document was then tested on a second set of subjects (n=6) each of whom classified 350 test images in four separate sequential batches using quantitative measures of precision and accuracy. The test set was sampled from 54,435 images from an additional 25 cameras. When compared to expert observers, they achieved a fair level of precision (Fleiss' kappa = 0.247) and 82.6% accuracy. Our approach can usefully be extended to other species and contexts.
Related
- A Parameter-Free Few-Shot Evaluation for Elephant Vocalisation Classification
- Simple but effective unsupervised classification for specified domain images: a case study on fungi images
- EEG-based classification models reveal differential neural processing of words and images
- Varietal classification and morphological traits prediction of lettuce cultivar (Lactuca sativa L.) based on RGB-images and convolutional neural network
- Comparative analysis of CNN, Transformer, and Mamba-based models for automated eye disease classification from fundus images
BibTeX
@misc{daecher2026training,
title = {{A training protocol for human classification of Asian elephant images from trail cameras}},
author = {F. Daecher and S. Venkatesh and A. Ayan and U. S. Weerathunga and T. V. Pushpakumara and V. P. M. K. Abeywardana and U. L. Thaufeek and S. de Silva},
howpublished = {bioRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.24.754213},
url = {https://www.biorxiv.org/content/10.64898/2026.09.24.754213v1},
note = {CC BY-NC-ND},
}
RIS
TY - UNPB
TI - A training protocol for human classification of Asian elephant images from trail cameras
AU - F. Daecher
AU - S. Venkatesh
AU - A. Ayan
AU - U. S. Weerathunga
AU - T. V. Pushpakumara
AU - V. P. M. K. Abeywardana
AU - U. L. Thaufeek
AU - S. de Silva
PY - 2026
DA - 2026/09/24
PB - bioRxiv
DO - 10.64898/2026.09.24.754213
UR - https://www.biorxiv.org/content/10.64898/2026.09.24.754213v1
AB - Trail cameras have become ubiquitous tools for ecological data collection over recent decades. Despite progress in the development of automated algorithms and artificial intelligence for image classification, our ability to process large volumes of data remain limited by the need for trained human observers to make refined judgements. We provide guidance on placement of trail cameras for observing Asian elephants (Elephas maximus) and outline a protocol for training and testing naive human observers in performing image classifications (age/sex class and group composition) that cannot yet be automated. This process can be used to develop a high-throughput workflow capable of extracting useful data from large volumes of images. Our training material consisted of 14,007 images collected from 6 trail cameras around Udawalawe National Park in Sri Lanka from 2017-2019. In the first stage, expert observers (n=3) trained a group of inexperienced participants (n=4), who engaged in an iterative process to develop a protocol document. The document was then tested on a second set of subjects (n=6) each of whom classified 350 test images in four separate sequential batches using quantitative measures of precision and accuracy. The test set was sampled from 54,435 images from an additional 25 cameras. When compared to expert observers, they achieved a fair level of precision (Fleiss' kappa = 0.247) and 82.6% accuracy. Our approach can usefully be extended to other species and contexts.
ER -