- Due to shifting economic and environment pressures on American cities, the amount of green spaces in cities is diminished when it is historically needed the most.
- In the face of climate change, cities need to strategically design green spaces to lower temperatures, improve air quality, and mitigate flooding.
- Using multiple sataelite images of google earth, identify patches of green space to train the model in style transfer.
- Specifically learning the color/shape/patterns of green spaces to then apply onto other map images that lack greenery.
- By generating greenery unto other cities that low green spaces density and/or high heat island effects. Approaches to AI would be through:
- Style Transfer: Tests the style of greener images being placed into content images of urban images
- Pix2Pix: Using manually applied colors on a map to achieve the appearance of more greenery