/ Computer Vision

CycleGAN : Image Translation with GAN (4)

Comments

Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks (CycleGAN) from UC Berkeley (pix2pix upgrade)

&

Learning to Discover Cross-Domain Relations with Generative Adversarial Networks (DiscoGAN) from SK T-Brain

DiscoGAN & CycleGAN

  • Almost Identical concept.
  • DiscoGAN came 15 days earlier. Low resolution ($64 \times 64$)
  • CycleGAN has better qualitative results ($256 \times 256$) and quantative experiments.

Difference from DTN

  • No $f$-constancy. Do not need pre-trained context encoder
  • Only need dataset $S$ and $T$ by proposing cycle-consistency

DiscoGAN

right fit

inline

without cross domain matching, GAN has mode collapse

inline

learn projection to mode in domain $B$, while two domains have one-to-one relation

Typical GAN issue: Mode collapse

inlineinline

  • top is ideal case, bottom is mode collapse failure case
  • Toy problem of 2-dim Gaussian mixture model
  • 5 modes of domain A to 10 modes of domain B

inline

  • GAN, GAN + const show injective mapping & mode collapse
  • DiscoGAN shows bijective mapping & generate all 10 modes of B.

proposed DiscoGAN

inline

CycleGAN has similar contribution on this point

inline
inline

Results

inline
inline

codes and more results in

CycleGAN

inline

Use more GAN techniques: LSGAN, use image buffer of previous generated samples

inline
inline
inline
inline
inline

failure case

inline

CycleGAN demonstrates more experiments!
project page : https://junyanz.github.io/CycleGAN/
code available with Torch and PyTorch

Junho Cho

Junho Cho

Integrated Ph.D course and Interested in Computer Vision, Deep Learning. For more information, tmmse.xyz/junhocho/

Read More