User trains DCGAN vision model on iPod touch 4 images

View on Reddit

A Reddit user trained a DCGAN from scratch using 350 images of a red solo cup taken with an iPod touch 4. The project explores training vision models on extremely small datasets and low-resolution hardware.