I have successfully automated the creation process for 4-panel comics using a combination of python scripts, imagemagick and (temporarily) calls to a free dog picture API.

Now I just need to train an image generator to generate images in a very specific style, and add this into the pipeline (replacing the current dog picture api thing) and hopefully then I'll have a comic generator.

This last part will take some time.

Reply to this note

Please Login to reply.

Discussion

I already have the images for the training, as I've been generating/creating one nearly every day for shibemag for 6 months and only selecting the ones I like. So, I have a pretty decent handle on the training data element.