What is this doc about?

Each of the detection dataset components (Single-Card, Simple-Bridge, Complex-Bridge)
has one or more subcomponents labeled by a three digit code (e.g. 003). This code
represents the unique id corresponding to the video from which those frames
were extracted. To load the frames on the bounding box annotator and annotate them,
we needed to scale them down. This doc specifies the scale-down parameter for each
set of frames.

If you ever expand the dataset, please keep track of any future scale factors.

-----
Components and Scale Factors:

simple_bridge/003: 0.4
simple_bridge/004: 0.2
single_card/005: 0.4
complex_bridge/006: 0.35
-----

Where to use these scale factors:
- line 170 in main.py in the Bounding box Annotator Code
- as the "scale" parameter at the crop_box.py executable 
  in the project Github repository