We provide a few tutorials to get you started with Rekall. We recommend starting with the Cyclist Detection Tutorial. This will walk you through Rekall's basic API and show you how to visualize Rekall queries with Vgrid.
To run these tutorials, you'll need to install vgrid and vgrid_jupyter for
visualization.
You should also clone the Rekall repository to get the tutorial notebooks:
git clone https://github.com/scanner-research/rekall
You'll need Python3.5 or greater.
pip3 install vgridpy
Or from source.
pip3 install vgrid_jupyter
jupyter-nbextension enable --py --sys-prefix vgrid_jupyter
Or from source.
Start with 01 Cyclist Detection.ipynb.
This will walk you through a simple example using person and bike detections to
detect a new class (bicyclists) using Rekall's operations.
Next we recommend moving on to 02 Empty Parking Space Detection.ipynb.
This tutorial will walk you through detecting empty parking spaces in a
static-camera feed of a parking lot using nothing more than the outputs of an
off-the-shelf object detector.
We recommend 03 Data Loading and Visualization.ipynb after completing
the other two tutorials.
This will walk you through how to load and visualize your own data using Rekall.
04 The Rekall Auto-Tuner.ipynb will teach you how to use Rekall's auto-tuner to
automatically tune the parameters of a query (the "magic numbers").