Evaluation and primer for using DeepLabCut to track small animals in large spaces
- 1University of Minnesota
Abstract
DeepLabCut (DLC), a machine learning-based pose estimation tool, has dramatically increased the throughput of behavioral tracking. Yet its implication remains limited by lack of validation data from diverse scenarios and users' unfamiliarity with the machine learning workflow. Here, we provide a practical primer on DLC terminology and workflow and performance evaluation of DLC in a novel scenario where the target is much smaller than those in previous validation data. We assessed model performance in a total of 36 model configurations across three training proportions, six training iterations, and two recording resolutions. Performance was evaluated on training convergence, prediction error, prediction certainty, and generalizability. Additionally, we selected an optimal model based on a balance between model performance and training time and applied it to 20 recordings at each resolution. In the test frames, 15 and 6 out of 18 model configurations met all performance criteria under high- and low-resolution, respectively. Of these, DLC was able to achieve low prediction error, high prediction confidence, and low train-test loss. Prediction error and confidence varied with training proportion, iteration, and resolution. Inference using the optimal models achieved similar high confidence predictions, with low-likelihood predictions predominantly associated with visual obstruction, low light, and reduced contrast between the subject and the background. Notably, the optimal models for high- and low-resolutions performed similarly despite a multi-fold difference in the cost of the recording devices, providing validation data for DLC in low-budget research, teaching and outreach scenarios.
Related
- Evaluation of the quantiplus rabies virus detection kit for diagnosis of rabies in humans and animals
- A fully immersive Virtual Reality setup using DeepLabCut reveals species-specific behavioural responses of freely moving coral reef fishes to conspecifics and predators
- Kinematic Analysis of Fine Motor Skills with the Montoya Staircase and Supervised Machine Learning: A framework from Video Acquisition to Behavioral Analysis Using DeepLabCut for Markerless Pose Estimation
- AmPair: automating housekeeping-gene primer design for species-level metataxonomics
- Advanced molecular poliovirus diagnostics across multiple real-time PCR platforms with locked nucleic acids and primer overhangs
BibTeX
@misc{kalra2026evaluation,
title = {{Evaluation and primer for using DeepLabCut to track small animals in large spaces}},
author = {L. Kalra and S. Balenger and S. P. Padhy and N. Lee and M. Xu},
howpublished = {bioRxiv},
year = {2026},
month = oct,
doi = {10.64898/2026.10.15.759451},
url = {https://www.biorxiv.org/content/10.64898/2026.10.15.759451v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Evaluation and primer for using DeepLabCut to track small animals in large spaces
AU - L. Kalra
AU - S. Balenger
AU - S. P. Padhy
AU - N. Lee
AU - M. Xu
PY - 2026
DA - 2026/10/15
PB - bioRxiv
DO - 10.64898/2026.10.15.759451
UR - https://www.biorxiv.org/content/10.64898/2026.10.15.759451v1
AB - DeepLabCut (DLC), a machine learning-based pose estimation tool, has dramatically increased the throughput of behavioral tracking. Yet its implication remains limited by lack of validation data from diverse scenarios and users' unfamiliarity with the machine learning workflow. Here, we provide a practical primer on DLC terminology and workflow and performance evaluation of DLC in a novel scenario where the target is much smaller than those in previous validation data. We assessed model performance in a total of 36 model configurations across three training proportions, six training iterations, and two recording resolutions. Performance was evaluated on training convergence, prediction error, prediction certainty, and generalizability. Additionally, we selected an optimal model based on a balance between model performance and training time and applied it to 20 recordings at each resolution. In the test frames, 15 and 6 out of 18 model configurations met all performance criteria under high- and low-resolution, respectively. Of these, DLC was able to achieve low prediction error, high prediction confidence, and low train-test loss. Prediction error and confidence varied with training proportion, iteration, and resolution. Inference using the optimal models achieved similar high confidence predictions, with low-likelihood predictions predominantly associated with visual obstruction, low light, and reduced contrast between the subject and the background. Notably, the optimal models for high- and low-resolutions performed similarly despite a multi-fold difference in the cost of the recording devices, providing validation data for DLC in low-budget research, teaching and outreach scenarios.
ER -