Skip to content

Latest commit

 

History

History
13 lines (8 loc) · 1002 Bytes

notebooks.md

File metadata and controls

13 lines (8 loc) · 1002 Bytes
Raw

Notebooks

We've organizated the workflow for prepping videos and running DLC with python code that is organized in notebooks. A notebook is an interactive interface for running code and a popular Python format for this is known as a 'Jupyter notebook' (file extenion '.ipynb'). We've stored jupyter notebooks for kineKit in the 'notebooks' directory within the root kineKit directory.

Here's a listing of the notebooks and their major functions:

  1. video_preprocess - For prepping video for DLC using videoTools. The code can handle image sequences or stand-alone video files and also allow for the selection of a region-of-interest. It is a very good idea to compress and downsample movies before using them to train DLC.

  2. dlc_run - Steps through an initial DLC training and acquisition of coordinates.

  3. dlc_evaluate - Evaluates the network, improves it, and uses it to analyze new videos.