Scaling up Deep Learning Based Super Resolution Algorithms Xiaoyong - - PowerPoint PPT Presentation

▶

Nov 01, 2022 217 likes •601 views

Scaling up Deep Learning Based Super Resolution Algorithms Xiaoyong Zhu Microsoft Cloud AI Group CNTK implementation Lets Enhance Image Source Image Source because human vision is more sensitive to luminance (black and white)

SLIDE 1

Scaling up Deep Learning Based Super Resolution Algorithms

Xiaoyong Zhu Microsoft Cloud AI Group

SLIDE 2

CNTK implementation

SLIDE 3

SLIDE 4

Let’s Enhance

SLIDE 5

SLIDE 6

Image Source

SLIDE 7

Image Source

SLIDE 8

because human vision is more sensitive to luminance (“black and white”) differences than chromatic differences

SLIDE 9

A few milestones including SRCNN, VDSR, DRRN, SRGAN

SLIDE 10

SRCNN (First to apply deep learning to Super Resolution, 2014)
VDSR (Very Deep Convolutional Networks, 2015)
DRRN (Deep Recursive Residual Network, CVPR 2017)
SRGAN (Photo-Realistic using GANs, CVPR 2017)
EDSR (Enhanced version using part of SRGAN’s work. Winner of

NTIRE2017 Super resolution challenge)

NTIRE Challenge (New Trends in Image Restoration and Enhancement) is a

challenge in this area (http://www.vision.ee.ethz.ch/ntire17/)

SLIDE 11

Image Source linear bilinear bicubic

SLIDE 12

SLIDE 13

Bicubic interpolation
VDSR (Very Deep Convolutional Networks, 2015)
DRRN (Deep Recursive Residual Network, CVPR 2017)
SRGAN (Photo-Realistic using GANs, CVPR 2017)
EDSR (Enhanced version using part of SRGAN’s work. Winner of

NTIRE2017 Super resolution challenge)

NTIRE Challenge (New Trends in Image Restoration and Enhancement) is a

challenge in this area (http://www.vision.ee.ethz.ch/ntire17/)

SLIDE 14

http://cs231n.github.io/understanding-cnn/

SLIDE 15

Link to paper

Bicubic SRCNN

SLIDE 16

SLIDE 17

Bicubic interpolation
SRCNN (First to apply deep learning to Super Resolution, 2014)
DRRN (Deep Recursive Residual Network, CVPR 2017)
SRGAN (Photo-Realistic using GANs, CVPR 2017)
EDSR (Enhanced version using part of SRGAN’s work. Winner of

NTIRE2017 Super resolution challenge)

NTIRE Challenge (New Trends in Image Restoration and Enhancement) is a

challenge in this area (http://www.vision.ee.ethz.ch/ntire17/)

SLIDE 18

ResNet architecture

Image frequency CNTK Code

SLIDE 19

Code available in CNTK

SLIDE 20

Bicubic interpolation
SRCNN (First to apply deep learning to Super Resolution, 2014)
VDSR (Very Deep Convolutional Networks, 2015)
SRGAN (Photo-Realistic using GANs, CVPR 2017)
EDSR (Enhanced version using part of SRGAN’s work. Winner of

NTIRE2017 Super resolution challenge)

NTIRE Challenge (New Trends in Image Restoration and Enhancement) is a

challenge in this area (http://www.vision.ee.ethz.ch/ntire17/)

SLIDE 21

SLIDE 22

SLIDE 23

SLIDE 24

The coolest idea in ML in the last twenty years - Yann Lecun

SLIDE 25

SLIDE 26

https://www.slideshare.net/xavigiro/deep-learning-for-computer-vision-generative-models-and-adversarial-training-upc-2016

z G(z) D(x) x D(G(z)) G D

SLIDE 27

http://people.eecs.berkeley .edu/~junyanz/projects/gvm/

SLIDE 28

SLIDE 29

Image source: http://kvfrans.com/visualizing-features-from-a- convolutional-neural-network/

SLIDE 30

SLIDE 31

https://www.cntk.ai/pythondocs/CNTK_302A_Evaluation_of_Pretrained_Su per-resolution_Models.html

Bicubic DRRN SRGAN

SLIDE 32

Bicubic DRRN SRGAN

https://www.cntk.ai/pythondocs/CNTK_302A_Evaluation_of_Pretrained_Su per-resolution_Models.html

SLIDE 33

SRCNN VDSR DRRN SRGAN EDSR

SLIDE 34

SLIDE 35

SLIDE 36

here here

SLIDE 37

Scalable Machine Learning using Kubernetes

Slides: bit.ly/DLwithK8S
Tutorial for deploying DL with K8S using acs_engine:

bit.ly/K8SwithACSEngine

Tutorial for deploying DL with managed K8S: aka.ms/AKS_GPU
Azure Machine Learning simplification to K8S: aka.ms/AMLtoACS
Batch AI for training DL at scale: bit.ly/deepbait

SLIDE 38