Recent Releases of deeplabcut_toolbox
deeplabcut_toolbox -
🐍 DeepLabCut Analysis Toolbox — Release v1.0.0
We are pleased to announce the initial public release of the DeepLabCut Analysis Toolbox (v1.0.0), a modular Python-based utility designed for post-processing of DeepLabCut (DLC) tracking data. This release establishes the toolbox as a standardized workflow for behavioral movement analysis in neuroscience research.
🧠 Overview
The toolbox provides: * automated velocity computation from DLC-tracked points, * detection of movement vs. non-movement periods using a customizable velocity threshold, * grouping of body parts for selective or combined analysis (e.g., head-only tracking in freezing behavior), * interval-based analysis over defined time windows (in frames), and * publication-ready visualizations of movement traces and periods of activity.
📁 Outputs
For each analyzed DLC .csv file, the toolbox automatically generates: * a .csv file with all computed velocities and movement flags, * one .pdf plot visualizing position, velocity, and movement indicators over time, * additional result files for each time interval if specified, and * a global .csv containing the average velocities across all DLC files in the dataset.
🛠️ User-defined Parameters
Users can configure the following key parameters: * pixelsize, timestep, movementthreshold (unit conversions and thresholding), * bodypartgroups (for grouped analyses), * time_intervals (for interval-based sub-analyses), * and path definitions for data input/output.
A detailed table of parameters and their descriptions is provided in the updated README.
📜 Citation
If you use this toolbox in your work, please cite it as:
Musacchio, Fabrizio. DeepLabCut Analysis Toolbox. Version 1.0.0. 2025. https://github.com/FabrizioMusacchio/MotilA
A CITATION.cff file is included for automated citation parsing. You may also use the Zenodo record once archived.
- Python
Published by FabrizioMusacchio about 1 year ago