Skip to content

README.md

Latest commit

 

History

History
15 lines (10 loc) · 527 Bytes

README.md

File metadata and controls

15 lines (10 loc) · 527 Bytes

StyleTransfer-VAE

Summary of the proposed VAE + Style Transfer Network

Purpose

The proposed VAE + Style Transfer Network is trained to be a generator, which could help us generate images with similar style compared with input samples and also the network could reconstruct the original images, unlike GAN which couldn't satisfy the reconstruction requirement.

Case Study I:

  • VAE + Style Transfer(file VAE+ST)

Case Study II:

  • VAE + Style Transfer + Supervised Learning (file VAE+ST+SL)