Skip to main content

Results & Discussion

Animation​

If you haven't seen this animation in our home page, this is the style transfer in action.

Comparison with others​

CycleGAN & DiscoGAN​

Result of CycleGAN & DiscoGAN

Overall the characters generated by CycleGAN are recognizable, however, some characters suffered from missing/broken strokes and some other appeared in the wrong direction.

DualGAN​

Result of DualGAN

For DualGAN, the structure of its generator and discriminator helps it to maintain a more stable and complete output structure. However, the strong constraints prevent it from learning the main characteristics of different fonts.

Ours​

Result of ours

The results from our model is visibily the best across all 4 fonts as we elimated previous issues that appeared in CycleGAN.

All models are trained for 100 epochs.

Style B​

Style C​

Style D​

Style E​

(Require UW Madison Login)
Style B Style C Style D Style E