Date: June 12, 2017
- Like I said, it’s all work in flux and progress, more soon.
- For a while now I’ve been experimenting with ways to use generative neural nets to make portraits.
- In most cases the quality is nowhere near the above, but in some cases the results are kinda interesting artistically.
- Specifically the second stage GAN is meta stable between smooth skin and hairy skin and often results in patchy output.
- Early experiments were based on deepdream-like approaches using backprop to the image but lately I’ve focused on GANs.
Related NLP Articles
Build smarter apps with our natural language processing API.