Hi,
I'm trying the Deep Dreamer demo, and I was wondering if there's a way to use a custom image database as training corpus, or if you had plans to include such features at some point.

(Also: if I process a video, how do I play the preview in the main window?)

Thanks,
Daniele