15
Privacy-Preserving Face Recognition Zekeriya Erkin1, Martin Franz2, Jorge Guajardo3, Stefan Katzenbeisser2, Inald Lagendijk1, and Tomas Toft4

Privacy-Preserving Face Recognition

  • Upload
    elisa

  • View
    81

  • Download
    4

Embed Size (px)

DESCRIPTION

Privacy-Preserving Face Recognition. Zekeriya Erkin1, Martin Franz2, Jorge Guajardo3, Stefan Katzenbeisser2, Inald Lagendijk1, and Tomas Toft4. Introduction. Alice. ?. Bob. Run face recognition to determine whether the face image is in database. Owns a face image - PowerPoint PPT Presentation

Citation preview

Page 1: Privacy-Preserving Face Recognition

Privacy-Preserving Face Recognition

Zekeriya Erkin1, Martin Franz2, Jorge Guajardo3,

Stefan Katzenbeisser2, Inald Lagendijk1, and Tomas Toft4

Page 2: Privacy-Preserving Face Recognition

Introduction

Alice Bob

Owns a face imageIs neither willing to share the image nor the detection result

Owns a face databaseIs not willing to reveal his data

Run face recognition to determine

whether the face image is in database?

Page 3: Privacy-Preserving Face Recognition

FIND

Page 4: Privacy-Preserving Face Recognition

Paillier cryptosystem

• additively homomorphic public-key encryption schemes

• [a + b] = [a][b],• [ab] = [a]b. ( b is a constant)

Page 5: Privacy-Preserving Face Recognition

Face Recognition

• Run Principal Component Analysis (PCA) from a set of criminal images to obtain eigenface

• Projects face images onto eigenfaces.

Page 6: Privacy-Preserving Face Recognition

PrincipalComponent Analysis (PCA)

• Θ1,Θ2, . . . , ΘM : vectors of length N• average of the training images• covariance matrix• Run PCA• To determine the face space, we selectK << M eigenvectors u1, . . . , uK that correspond to the K largest eigenvalues.

Page 7: Privacy-Preserving Face Recognition

criminal projection

• Θ1,Θ2, . . . , ΘM are projected onto the subspace spanned by the basis u1, . . . , uK to obtain their feature vector representation Ω1, . . . , ΩM.

Page 8: Privacy-Preserving Face Recognition

Suspect Projection

• input image Γ

Page 9: Privacy-Preserving Face Recognition

Progection in the encrypted domain

• input image

Page 10: Privacy-Preserving Face Recognition

Calculating distances

Client’s

Page 11: Privacy-Preserving Face Recognition

Calculating distances(conti)

BobAlice

Page 12: Privacy-Preserving Face Recognition

Calculating distances

Page 13: Privacy-Preserving Face Recognition

•DEMO~

Page 14: Privacy-Preserving Face Recognition

Paillier cryptosystem

• Two large prime number p, q• n = p*q• Select random integer g • •

• Encryption• Decryption

Private key

Public key

Page 15: Privacy-Preserving Face Recognition

Paillier cryptosystem

• Homomorphic addition of plaintexts

• Homomorphic multiplication of plaintexts