Variational Auto Encoder入門

Copyright (C) 2019 DeNA Co.,Ltd. All Rights Reserved. n
n Unsupervised feature learning with deep generative model Variational Auto-Encoder Adversarial feature learning n 2

+> TS ( ) U Twitter: @0hnishi , speakerdeck: https://speakerdeck.com/katsunoriohnishi n 6P U 2014,4/-2017,9/: B4~M2.509+7Computer VisionGH KAM • 3L (8QC@): CVPR16 (spotlight oral), ACMMM16, AAAI18 (oral) U RN: http://katsunoriohnishi.github.io/ U 2017,10/-<B: DeNA AI!%)1 • DEDeNA(!#"$):4Computer Vision/Data Science?5 '* & 2.= U → https://dena.ai/work7/ U JI-;OF 3

26 +7∩deep learning∩&.# $3951.!" 4 (-4%,*() 26 1. 395: >< Kaggle8)% VAE/'0 0

8<0=∩deep learning∩+3'$(9?;73%"!#& 5 ( 2:*1/-) 8<$ 73 9?; @>< Kaggle>.* VAE5,6 6 9)4

Copyright (C) 2019 DeNA Co.,Ltd. All Rights Reserved. Auto-encoder n
encoder-decoder &" 6 $' loss ! Mean Squared Error(MSE) Self-supervised learning % # https://hackernoon.com/a-deep-convolutional-denoising-autoencoder-for-image-classification-26c777d3b88e

Copyright (C) 2019 DeNA Co.,Ltd. All Rights Reserved. Auto-encoder
7 Output!mask segmentation “Auto” encoder-decoder'%)MNA> Segmentation Auto-encoder.48O encoder!=:K *G(&) %$→CG(&) "#3 Pre-training MSE<P6BV!1; 8O NoiseF BV!2 noiseE R27H?I/noise0 Noise & +5T27%$!LDJ@?QU Bottleneck layer!9, -S6 [V. Badrinarayanan+, CVPR15] https://towardsdatascience.com/applied-deep-learning-part-3-autoencoders-1c083af4d798 [X. Zhan+, AAAI18]

Copyright (C) 2019 DeNA Co.,Ltd. All Rights Reserved. Auto-encoder n
Pre-training: C 7%59 C Test $∩ train … 5> n Noise=; 0 n Anomaly detection18.?*6 n &(A<3)/',)!+/ C MSE 4B#8. "@&-2! 8

Copyright (C) 2019 DeNA Co.,Ltd. All Rights Reserved. Variational Auto-encoder
9 Auto-encoder Variational Auto-encoder

$("!#.*"!#- 2 1)encoder-decoder$("!# +0 $( %,&.*"!# '/"!# 10 https://stanford.edu/~shervine/teaching/cs-229/cheatsheet-supervised-learning

VAE ( #$'! ( & z %" 11

n 12 https://qiita.com/kenmatsu4/i tems/b029d697e9995d93aa24

n 5,-#"3 14 /+45,0. *! +4&!1' 5,0( +4 2 $%)1' https://hackernoon.com/latent-space-visualization-deep-learning-bits-2-bd09a46920df

n Q:VAE#!=-0$2":6 ? A: ' 15 ()/":) &4":713.5>+": VAE#!=-.*→output;*9,"713.5 8%< )/":( Tutorial on Variational Autoencoders [C. Doersch, arxiv15]

n predict/train 16

n Forward" ! 17 ! " Σ " z Input X output Y Sample $ from % ! " , Σ " z # backward

n 18 ! " Σ " z Input X output Y Sample $ from % Ο, Ι ∗ + Forward Backward

n Loss 19 ! " Σ " z Input X output Y Sample $ from % Ο, Ι ∗ + + − " - ./ %(!("), 2("))||% Ο, Ι KL !* q(z|X) &X $,6'logp(X|z) +/0 − (%&, *KL ! -.#2 1'4 *5') "30

n 7459JG;? L Q: -/'.$+'!# … L A: MSE=E! 20 DF,/0'> H3<9β" KL)$,&%0' q(z|X) 5 *(X 29K6logp(X|z) 8@C − 7459 KL)$,&%0' A:IB (e.g. MSE) c β # − $ % A:*(:*( 6 →1 1: *( 8@

n ! / • https://arxiv.org/pdf/1606.05908.pdf • https://qiita.com/kenmatsu4/items/b029d697e9995d93aa24 21 −

n : 23 ! " Σ " z Input X output Y Sample $ from % Ο, Ι ∗ + c

n +36 AEVAE'. 7 5,-)$1/% 24 AE!4 5,-)$1 #0*2&( " VAE https://qiita.com/kenmatsu4/i tems/b029d697e9995d93aa24

n :>4 B */DL&%&'15example- • Chainer: https://github.com/chainer/chainer/tree/master/examples/vae • Pytorch: https://github.com/pytorch/examples/tree/master/vae • Keras: https://github.com/keras-team/keras/blob/master/examples/variational_autoencoder.py B Tips • " !)+,=3< • A67.(0? #$&2 • 0?@9 8; 25

n $,(+#&2018%*"&(*%)"'%A! … K −"#$ −β 26 pytorch [Kingma+, ICLR14] J=?/ 7>-G: 9;H@ :;H .6 CB< F https://github.com/pytorch/examples/blob/master/vae/main.py e.g.) D9IE 543

Copyright (C) 2019 DeNA Co.,Ltd. All Rights Reserved. GAN n
VAE MSE Loss GAN 27 https://skymind.ai/wiki/generative-adversarial-network-gan

Copyright (C) 2019 DeNA Co.,Ltd. All Rights Reserved. GAN n
Discriminator#+1,3/ )- 5 "$( %' 28 real/fake0* discriminator !24,3/& . ,3/ UCF101 Chance 0.9 unsupervised VGAN [C. Vondrick+, NIPS16] 36.7 FTGAN [K. Ohnishi+, AAAI18] 60.9 supervised Two-stream [K. Simonyan+, NIPS14] 88.0 I3D [J. Carreira+ CVPR17] 98.0 CIFAR10 Chance 10.0 unsupervised DCGAN [A. Radford+, ICLR16] 82.8 supervised Alexnet [A. Krizhevsky+, NIPS12] 89.0 Resnet110 [K. He+, CVPR16] 93.6 Accuracy

learning [J. Donahue+ ICLR17] n GeneratorEncoderadversarial training G: z (uniform distribution) G(z) E: x E(x) 30 generated data e.g.) uniform distribution BiGAN

learning [J. Donahue+ ICLR17] n GeneratorEncoderadversarial training G: z (uniform distribution) G(z) E: x E(x) 31 real data generated data e.g.) uniform distribution BiGAN

learning [J. Donahue+ ICLR17] n GeneratorEncoderadversarial training G: z (uniform distribution) G(z) E: x E(x) 32 real data generated data e.g.) uniform distribution generated feature BiGAN

learning [J. Donahue+ ICLR17] n 33 GAN BiGAN xG(z) Dreal fake • x: real data • G(z): {x, E(x)}{G(z), z} D • x: real data • E(x): real data • z: random noise • G(z) :

learning [J. Donahue+ ICLR17] n 34

learning [J. Donahue+ ICLR17] n 35

learning [J. Donahue+ ICLR17] n : https://github.com/jeffdonahue/bigan pytorch( ): https://github.com/9310gaurav/ali-pytorch 36

Copyright (C) 2019 DeNA Co.,Ltd. All Rights Reserved. VAEBiGAN
37 n VAE > -:7<+0* > #'(23 61 > Encoder, Decoder$! "%7<+# > )89 n BiGAN > -:.0* > #'(23 61 > Encoder, Generator, Discriminator$! "%/61 > 5;=&,+GAN 4

VAE;M<C N 2:/= =*7 n GANBKD?5+ELF ↑AI0J∩deep learning∩.5'%(BKD?5>E LF n -13HEnd2End$#"64,9!&8G ) @ 38

Copyright (C) 2019 DeNA Co.,Ltd. All Rights Reserved. FAQ n
$#LGDO93;31Ca!S"'*)_W9 3;' b _W * IFXK_W93;E @`F_WUH? b IFYVC >B n Kaggle 0=8VAEVNU? J^) b \ #AENU? )'VAEV &(NU? )T+*) n VAE! A - b !524</Z] >B31#,PQ!% b AE,VAE)&( R b 6.7:#! M[!% 39

Variational Auto Encoder入門

Variational Auto Encoder入門

More Decks by Katsunori Ohnishi

Other Decks in Technology

Featured

Transcript