Contenu connexe Similaire à Wasserstein GAN Tfug2017 07-12 (8) Plus de Yuta Kashino (20) Wasserstein GAN Tfug2017 07-122. Yuta Kashino ( )
BakFoo, Inc. CEO
Astro Physics /Observational Cosmology
Zope / Python
Realtime Data Platform for Enterprise / Prototyping
3. Yuta Kashino ( )
arXiv
stat.ML, stat.TH, cs.CV, cs.CL, cs.LG
math-ph, astro-ph
- PyCon2016
- PyCon2017 Edward
- 2017 8 TFUG
@yutakashino
https://www.slideshare.net/yutakashino/pyconjp2016
7. GAN 1
- Generative Adversarial Networks
- Ian Goodfelow
- Bengio , Theano/Pyleran2
- Google Brain
- 2016 NIPIS Tutorial
- : The GAN Zoo
https://goo.gl/uC8xn2
https://github.com/hindupuravinash/the-gan-zoo
8. GAN 2
- GAN …
- Meow Generator
- HDCGAN, WGAN, LSGAN…
https://ajolicoeur.wordpress.com/cats/
https://github.com/hindupuravinash/the-gan-zoo
10. DCGAN
-
- CNN
- G /D
- Pool/ Full Batch Norm, Leaky ReLU
Unsupervised Representation Learning with
Deep Convolutional Generative Adversarial Networks
https://arxiv.org/abs/1511.06434
https://goo.gl/8EmZgT
14. WGAN 2
-
- Read-through: Wasserstein GAN
- Wasserstein GAN and the Kantorovich-Rubinstein
Duality
-
https://goo.gl/7ywVwc
https://goo.gl/40eCbR
18. 4
- Total Variation(TV)
- Kullback-Leibler (KL) divergence
- Jenson-Shannon (JS) divergence
- Earth Mover (EM) / Wasserstein
(Pr, Pg) = sup
A
|Pr(A) Pg(A)|
KL(PrkPg) =
Z
x
log
✓
Pr(x)
Pg(x)
◆
Pr(x) dx
JS(Pr, Pg) =
1
2
KL(PrkPm) +
1
2
KL(PgkPm)
M = Pr/2 + Pg/2M
W(Pr, Pg) = inf
2⇧(Pr,Pg)
E(x,y)⇠
⇥
kx yk
⇤
19. 4
- W
- JS
- KL
- TV
KL(P0kP✓) = KL(P✓kP0) =
(
+1 if ✓ 6= 0 ,
0 if ✓ = 0 ,
(P0, P✓) =
(
1 if ✓ 6= 0 ,
0 if ✓ = 0 .
JS(P0, P✓) =
(
log 2 if ✓ 6= 0 ,
0 if ✓ = 0 ,
W(P0, P✓) = |✓|
U[0, 1]
https://goo.gl/40eCbR
21. 3 2
3: Kantorovich-Rubinstein
- W
- W
max
w2W
Ex⇠Pr
[fw(x)] Ex⇠P✓
[fw(x)] sup
kfkLK
Ex⇠Pr
[f(x)] Ex⇠P✓
[f(x)]
= K · W(Pr, P✓)
r✓W(Pr, P✓) = r✓(Ex⇠Pr
[fw(x)] Ez⇠Z[fw(g✓(z))])
= Ez⇠Z[r✓fw(g✓(z))]
22. W/EM 1
-
-
W(Pr, Pg) = inf
2⇧(Pr,Pg)
E(x,y)⇠
⇥
kx yk
⇤
scypy.optimize.linprog
γ
https://goo.gl/7ywVwc
https://goo.gl/7ywVwc
24. 3 2( )
3: Kantorovich-Rubinstein
- W
- W
max
w2W
Ex⇠Pr
[fw(x)] Ex⇠P✓
[fw(x)] sup
kfkLK
Ex⇠Pr
[f(x)] Ex⇠P✓
[f(x)]
= K · W(Pr, P✓)
r✓W(Pr, P✓) = r✓(Ex⇠Pr
[fw(x)] Ez⇠Z[fw(g✓(z))])
= Ez⇠Z[r✓fw(g✓(z))]
26. r✓W(Pr, P✓) = r✓(Ex⇠Pr
[fw(x)] Ez⇠Z[fw(g✓(z))])
= Ez⇠Z[r✓fw(g✓(z))]
35. -
-
- WGAN
- GAN G D
Improved Training of Wasserstein GANs
https://arxiv.org/abs/1704.00028
Do GANs actually learn the distribution? An empirical study
https://arxiv.org/abs/1706.08224