Generate new Whataburger store concepts the likes of which have never been seen!
WhataGAN is an exercise in data collection, organization, refinement, and synthesis.
After installing and activating the conda environment run:
python whatagan
Metadata will be stored in JSON format with the following schema:
{
//
"location": "Whataburger+2424+Baldwin+Blvd,Corpus-Christi,TX",
// Internal identifier (will be replaced as key by store)
"uuid": "57e00294bce4451ebf3807747200b911",
// Whataburger Store #
"number": 2,
// Can StreetView see this location?
"present": false,
// GAN needs at least 512x512
"size": "600x600",
// Orientation range of camera where store can be seen
"heading": [
0,
340
],
// Zoom range of camera where store can be seen
"fov": [
80,
81
],
// Pitch range of camera where store can be seen
"pitch": [
10,
11
]
}
- Incorporate MongoDB for persistent metadata storage
- Add Streamlit UI for easy metadata manipulation