A deep learning framework for multi-animal pose tracking.
-
Updated
Sep 9, 2026 - Python
A deep learning framework for multi-animal pose tracking.
Behavioral segmentation of open field in DeepLabCut, or B-SOID ("B-side"), is a pipeline that pairs unsupervised pattern recognition with supervised classification to achieve fast predictions of behaviors that are not predefined by users.
Sample datasets for SLEAP
A package for extracting facial expressions from SLEAP analyses
Desktop SLEAP workflow manager for local labeling and Great Lakes HPC training/inference with Slurm, SSH, and task history automation.
Jupyter notebooks for running SLEAP on Google Colaboratory
Data pipelines for a mouse behavior lab: video ingest to network storage, SLEAP model training and evaluation, and multi-GPU pose inference, run in production across multiple cohorts. Plus per-cohort video cropping for a three-chamber assay.
Local, pose-based, human-in-the-loop behavior classifier: a lightweight modern JAABA (FastAPI + sleap-io + movement + scikit-learn)
To associate your repository with the sleap topic, visit your repo's landing page and select "manage topics."