Hello everyone,

Thanks for your awesome work on Krita. I want to highlight an idea. Recently deep-learning was used to generate a new image from simple.
It can be used to start sketches/drawing. I will give some examples of this technology which will be very helpful for the artist to have this tool at Krita.

1. Image2Image https://affinelayer.com/pixsrv/
2. HyperRealistic Face Generator https://medium.com/syncedreview/gan-2-0 ... 439d33ebaf
3. Landscape Generator https://www.theverge.com/2019/3/19/1827 ... landscapes

There are other ideas also like super-resolution, day to night, filters, depth estimation.

Please consider using this technology and I recommend using Dlib library with pre-trained models.

Regards