BehaveTrack uses user-specific B-SOiD/A-SOiD classifier to predict behavior based on pose. Generate ethogram and summary statistics for animal recordings. Intuitive interface for tagging and grouping files. Enhance animal behavior studies with accurate and reliable predictions.

Install

create a conda environment

conda create -n behavetrack python==3.9 -y 
conda activate behavetrack

clone github repo and change directory into repo

git clone https://github.com/runninghsus/BehaveTrack.git
cd BehaveTrack

install dependencies

pip install -r requirements.txt

Usage

Run the streamlit app

streamlit run behavetrack.py

Existing

Quality of life improvements:

• Drag and drop B-SOiD trained classifier
• Drag and drop A-SOiD trained classifier
• Using example video to annotate user-definition
• Organize files into 2+ conditions for comparisons, including the following:

Reactive post-hoc analyses:

• ethogram
• behavior location
• behavior ratio
• behavioral frequency
• behavior duration
• behavior transition

In progress

Quality of life improvements:

• annotation retrieval from A-SOiD

Reactive post-hoc analyses:

• behavioral pose speed

Project funding

Glen de Vries Presidential Fellowship for Biological Sciences at Carnegie Mellon University.

Website

This is currently developed by Alex Hsu to support research using B-SOiD.

LICENSE

BehaveTrack is released under a BSD 3-Clause "New" or "Revised" License and is intended for research/academic use only.

---