Artistic Stylization of Face Photos Based on a Single Exemplar - - PowerPoint PPT Presentation

artistic stylization of face photos
SMART_READER_LITE
LIVE PREVIEW

Artistic Stylization of Face Photos Based on a Single Exemplar - - PowerPoint PPT Presentation

Zili Yi Supervisor: Minglun Gong 2015-11-01 Artistic Stylization of Face Photos Based on a Single Exemplar Outline Introduction Background Method Results Conclusion Reference input Introduction example Style Transfer


slide-1
SLIDE 1

Artistic Stylization of Face Photos Based on a Single Exemplar

Zili Yi Supervisor: Minglun Gong 2015-11-01

slide-2
SLIDE 2

Outline

 Introduction  Background  Method  Results  Conclusion  Reference

slide-3
SLIDE 3

Introduction

input example

  • utput

Style Transfer Stylization of Faces Non-photorealistic Rendering

Challenge

Region-specific Viewers’ Sensitivity to face Variation of faces

slide-4
SLIDE 4

Style-specific methods

Pencil drawing, Lu c. et. al. Tessellation, Li h. et. Al. Halftoning, Li H. et. al. Screening, Li H. et. Al.

slide-5
SLIDE 5

Data driven methods

Guided Transfer, Li H. Style transfer, Shih Y. et. al. Data Driven Cartoonizing, Zhang

  • Y. et. al.
slide-6
SLIDE 6

Our Task Data Driven Style-Specific General stylization with a single Example

slide-7
SLIDE 7

Background

 Face Detection & Alignment  Non-Photo Realistic Rendering  Color Style Transfer  Optimized Texture Synthesis

slide-8
SLIDE 8

Framework of Our Method

Example Color Transfer Texture Transfer

slide-9
SLIDE 9

Framework of Our Method

Example Color Transfer Texture Transfer

slide-10
SLIDE 10

Framework of Our Method

Example Color Transfer Texture Transfer

slide-11
SLIDE 11

Color Transfer

Example Color Transfer Texture Transfer

slide-12
SLIDE 12

Color Transfer

Face landmark

label assignment dense correspondence color assignment

face detection& normalization 320*448 image

slide-13
SLIDE 13

Texture Transfer

Example Color Transfer Texture Transfer

slide-14
SLIDE 14

Texture Transfer

 We view the problem of Texture Transfer as Edge-

Preserved Texture synthesis problem

 Optimized Texture synthesis

 Iteratively optimizing and voting

slide-15
SLIDE 15

Optimizing

 Finding the Nearest Neighbor Field  Generalized PatchMatch

‘ measures the L2 distance Between patch and S E

slide-16
SLIDE 16

Voting

 Simply averaging the Nearest Neighbors

slide-17
SLIDE 17

Pyramid

 The texture is synthesized in a pyramid manner

 Coarse to fine  10 scales (ratio 1.5)

slide-18
SLIDE 18

Preserve the edges

 Pyramid Edge Map

 Using edge maps as the preserving mask

slide-19
SLIDE 19

The final synthesized image

slide-20
SLIDE 20

Experiment Result

slide-21
SLIDE 21

Results

slide-22
SLIDE 22

Comparison

slide-23
SLIDE 23

Comparison

slide-24
SLIDE 24

Conclusion

 Strength

 Robust to gender, skin color, hair style, face accessories, beards,

glasses, slight variation in poses, lightening condition

 Able to transfer wide varieties of styles (pencil drawing, sand

drawing, oil painting, mosaic, screening, water color painting, Chinese painting and Pyrography)

 Limitation

 Time consumption (over 40 mins to process one image)  Unable to handle stylization which involves geometry

exaggeration

 T

end to blur textures with long curve patterns.

slide-25
SLIDE 25

Reference

slide-26
SLIDE 26

Reference