StyledFontGAN. Generate a complete alphabet from a single letter using AI.

github.com/joshpc/StyledFontGAN

Problem

The full read on this project is a sign-in away — what it solves, who it serves, and where it sits in the market.

Audience

The full read on this project is a sign-in away — what it solves, who it serves, and where it sits in the market.

Market

The full read on this project is a sign-in away — what it solves, who it serves, and where it sits in the market.

Pricing

The full read on this project is a sign-in away — what it solves, who it serves, and where it sits in the market.

Vaya's read on this project

Problem, audience, market, and the verdict — sign in to see it.

Updates

April 2020

Removes old notes Apr 30
Updates README Apr 24
Updates results with the proper l1 loss + adversarial Apr 24
Updates playground to use l1 + adversarial loss Apr 23
Cleans up the playground a little Apr 23
Updates README with better references Apr 23
Updates README Apr 22
Attempts with larger batch sizes and dimensions Apr 20
Uploads results Apr 14
Updated the Playground to show results a little better Apr 14
Additional refactors Apr 14
Fixes issue in Playground that caused errors between different runs o… Apr 13
Uses L1 loss for generator, and normal wgan style discriminator Apr 13
Refactors training and loss Apr 13
Adds second training method (generator only, with L1 loss) as a simpl… Apr 8
Adds additional CONV layers and more hyperparameters to tweak Apr 6
Merge pull request #1 from joshpc/add-skeleton Apr 1
Adds environment and loss visualization framework Apr 1

March 2020

Generate data and pass it through a discriminator Mar 30
Adds a simple Conv layer (1 -> N) channels to make the model work Mar 27