Let's talk about Fakes - Speaker Deck

Slide 1

Slide 1 text

Let's talk about Fakes Let's talk about all the good things And the bad things that may be

Slide 2

Slide 2 text

Making Faces • Face Synthesis  from nothing to face • Facial Attributes   modify age, gender, add glasses change hairstyle or color • Face Swap  use the body and background of target, face of source • Facial Expression Transfer  transfer only the expression or modify expression to match an impersonator or an audio track (reenactment)

Slide 3

Slide 3 text

No content

Slide 4

Slide 4 text

Face Synthesis https://research.nvidia.com/publication/2017-10_Progressive-Growing-of

Slide 5

Slide 5 text

https://thispersondoesnotexist.com/

Slide 6

Slide 6 text

Facial Attributes

Slide 7

Slide 7 text

https://www.faceapp.com/

Slide 8

Slide 8 text

https://www.faceapp.com/

Slide 9

Slide 9 text

Face Off (1997)

Slide 10

Slide 10 text

Nicolas Cage Everywhere

Slide 11

Slide 11 text

Reenactment Cafe 80s, Hill Valley, USA, (October 21, 2015)

Slide 12

Slide 12 text

Jordan Peele (2018)

Slide 13

Slide 13 text

Face2Face (2016)

Slide 14

Slide 14 text

RGB-D tracking

Slide 15

Slide 15 text

No content

Slide 16

Slide 16 text

No content

Slide 17

Slide 17 text

FaceSwapGAN (2018)

Slide 18

Slide 18 text

No content

Slide 19

Slide 19 text

No content

Slide 20

Slide 20 text

No content

Slide 21

Slide 21 text

No content

Slide 22

Slide 22 text

No content

Slide 23

Slide 23 text

No content

Slide 24

Slide 24 text

No content

Slide 25

Slide 25 text

No content

Slide 26

Slide 26 text

FaceShifter (Dec 2019)

Slide 27

Slide 27 text

No content

Slide 28

Slide 28 text

No content

Slide 29

Slide 29 text

No content

Slide 30

Slide 30 text

How to Spot a Fake

Slide 31

Slide 31 text

https://www.kaggle.com/c/deepfake-detection-challenge

Slide 32

Slide 32 text

uneven skin tone artifacts in hair says strange things warped edges How to Spot a Fake obstruction glitches

Slide 33

Slide 33 text

That was a real photo.

Slide 34

Slide 34 text

uneven skin tone artifacts in hair says strange things warped edges How to Spot a Fake obstruction glitches

Slide 35

Slide 35 text

https://people.csail.mit.edu/mrub/vidmag/

Slide 36

Slide 36 text

Why this is important

Slide 37

Slide 37 text

We are hardwired to react to faces.

Slide 38

Slide 38 text

No content

Slide 39

Slide 39 text

Especially to moving faces.   Even more so, emotional.

Slide 40

Slide 40 text

No content

Slide 41

Slide 41 text

We have a tendency to fall for and spread shocking images.

Slide 42

Slide 42 text

No content

Slide 43

Slide 43 text

No content

Slide 44

Slide 44 text

We react to images that speak to our tribal instincts or align with our political world view

Slide 45

Slide 45 text

No content

Slide 46

Slide 46 text

No content

Slide 47

Slide 47 text

No matter how we try, how many tools we create to detect fakes, we WILL BE AFFECTED...