[pr12] intro. to gans jaejun yoo

GANsPR12와 함께 이해하는

* Generative Adversarial Nets Ian Goodfellow et al. 2014를 바탕으로 작성한 리뷰

Jaejun YooPh.D. Candidate @KAIST

PR12

16th Apr, 2017

안녕하세요 저는

유재준

- Ph.D. Candidate

- Medical Image Reconstruction,

- http://jaejunyoo.blogspot.com/

Topological Data Analysis, EEG

http://jaejunyoo.blogspot.com/

Generative Adversarial Network

PREREQUISITESGenerative Models

“FACE IMAGES”


* Figure adopted from BEGAN paper released at 31. Mar. 2017 David Berthelot et al. Google (link)

Generated Images by Neural Network

https://128.84.21.199/pdf/1703.10717.pdf


“What I cannot create, I do not understand”


“What I cannot create, I do not understand”

If the network can learn how to draw cat and dog separately, it must be able to classify them, i.e. feature learning follows naturally.

PREREQUISITESTaxonomy of Machine Learning

From Yann Lecun, (NIPS 2016)From David silver, Reinforcement learning (UCL course on RL, 2015)

PREREQUISITES

Slide adopted from Namju Kim, Kakao brain (SlideShare, AI Forum, 2017)

y = f(x)

PREREQUISITES


PREREQUISITESTaxonomy of Machine Learning

From Yann Lecun, (NIPS 2016)From David silver, Reinforcement learning (UCL course on RL, 2015)

PREREQUISITES


PREREQUISITES


* Figure adopted from NIPS 2016 Tutorial: GAN paper, Ian Goodfellow 2016

Generative Adversarial Network

SCHEMATIC OVERVIEW

z

G

D

x

Real or Fake?

Diagram ofStandard GAN

Gaussian noise as an input for G

z

G

D

x

Real or Fake?


지폐위조범

경찰

SCHEMATIC OVERVIEW

z

G

D

x

Real or Fake?


지폐위조범

경찰

QP

SCHEMATIC OVERVIEW


Data distribution Model distributionDiscriminator

SCHEMATIC OVERVIEW

* Figure adopted from Generative Adversarial Nets, Ian Goodfellow et al. 2014

Minimax problem of GAN

THEORETICAL RESULTS

Show that…

1. The minimax problem of GAN has a global optimum at 𝒑𝒑𝒈𝒈 = 𝒑𝒑𝒅𝒅𝒅𝒅𝒅𝒅𝒅𝒅

2. The proposed algorithm can find that global optimum

TWO STEP APPROACH

THEORETICAL RESULTSProposition 1.

THEORETICAL RESULTSMain Theorem

THEORETICAL RESULTSConvergence of the proposed algorithm

THEORETICAL RESULTSConvergence of the proposed algorithm

"The subderivatives of a supremum of convex functions include the derivative of the function at the point where the maximum is attained."

RESULTS

* Figure adopted from DCGAN, Alec Radford et al. 2016 (link)

What can GAN do?

https://arxiv.org/pdf/1511.06434.pdf

RESULTS

What can GAN do?

Vector arithmetic(e.g. word2vec)

RESULTS

“We want to get a disentangled representation space EXPLICITLY.”

Neural network understanding “Rotation”



DIFFICULTIES

DIFFICULTIES CONVERGENCE OF THE MODEL

DIFFICULTIES HOW TO EVALUATE THE QUALITY?

DIFFICULTIES MODE COLLAPSE (SAMPLE DIVERSITY)

* Slide adopted from NIPS 2016 Tutorial, Ian Goodfellow

RELATED WORKS

* Unrolled GAN Luke Metz et al. 2016

RELATED WORKS

* CycleGAN Jun-Yan Zhu et al. 2017

* SRGAN Christian Ledwig et al. 2017

Super-resolution

Img2Img Translation

RELATED WORKS

* infoGAN Xi Chen et al. 2016

Find a CODE

RELATED WORKS

Find a CODE

* infoGAN Xi Chen et al. 2016

RELATED WORKS

“The information in the latent code c should not be lost in the generation process.”

c

z

G

D

x

I

Real or Fake?

Mutual Info. infoGAN: maximize I(c,G(z,c))

Diagram ofinfoGAN Impose an extra constraint to learn disentangled feature space

THANK YOU [email protected]

[pr12] intro. to gans jaejun yoo

Technology