SCiFI - A System for Secure Face Identification

@article{Osadchy2010SCiFIA,
  title={SCiFI - A System for Secure Face Identification},
  author={Margarita Osadchy and Benny Pinkas and Ayman Jarrous and Boaz Moskovich},
  journal={2010 IEEE Symposium on Security and Privacy},
  year={2010},
  pages={239-254}
}
We introduce SCiFI, a system for Secure Computation of Face Identification. The system performs face identification which compares faces of subjects with a database of registered faces. The identification is done in a secure way which protects both the privacy of the subjects and the confidentiality of the database. A specific application of SCiFI is reducing the privacy impact of camera based surveillance. In that scenario, SCiFI would be used in a setting which contains a server which has a… 

Figures and Tables from this paper

Reconstructing a Fragmented Face from an Attacked Secure Identication Protocol
TLDR
This thesis develops a novel approach to initially assemble the information given by the SCiFI protocol to create a fragmented face and explores novel methods of comparing images from different subspaces using metric learning and other forms of facial descriptors.
Reconstructing a fragmented face from a cryptographic identification protocol
TLDR
The consequences of malformed input attacks on SCiFI are studied - from both a security and computer vision standpoint - to underscore the risk posed by malicious adversaries to todays automatic face recognition systems.
SECURE COMPUTATION OF FACE IDENTIFICATION-SCiFI
TLDR
This work studies the consequence of malformed input attacks on secure face recognition systems-from both a security and computer vision standpoint, and proposes a visualization approach that exploits this breach.
Privacy Preserving Face Identification in the Cloud through Sparse Representation
TLDR
This is the first work that introduces sparse representation to the secure protocol of private face identification, which reduces the dimension of the face representation vector and avoid the patch based attack of a previous work.
SECURE COMPUTATION OF FACE IDENTIFICATION - SCiFI
TLDR
This work studies the consequence of malformed input attacks on the secure face recognition system-from both a security and computer vision standpoint, and proposes a visualization approach that exploits this breach.
A Secure Face-Verification Scheme Based on Homomorphic Encryption and Deep Neural Networks
TLDR
Face features are extracted using deep neural networks and then encrypted with the Paillier algorithm and saved in a data set and show that the approach can enhance the security of a recognition system with little decrease in accuracy.
It is All in the System's Parameters: Privacy and Security Issues in Transforming Biometric Raw Data into Binary Strings
TLDR
Many of the existing “privacy preserving” solutions neglect the privacy and security aspects of the feature extraction and binarization processes, and it is urged to close this gap in the security and privacy of biometric systems.
Efficient blind face recognition in the cloud
TLDR
The experimental results show that the efficiency of the two schemes is greatly improved compared with SCiFI schemes, and the second scheme improves recognition accuracy greatly.
Fully Private Noninteractive Face Verification
TLDR
This work presents a private face verification system that can be executed in the server without interaction, working with encrypted feature vectors for both the templates and the probe face, and opens the door to completely private and noninteractive outsourcing of face verification.
Fully Private Non-interactive Face Verification
TLDR
This work presents a private face verification system that can be executed in the server without interaction, working with encrypted feature vectors for both the templates and the probe face, and opens the door to completely private and non-interactive outsourcing of face verification.
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 36 REFERENCES
Privacy-Preserving Face Recognition
TLDR
This paper proposes for the first time a strongly privacy-enhanced face recognition system, which allows to efficiently hide both the biometrics and the result from the server that performs the matching operation, by using techniques from secure multiparty computation.
Preserving privacy by de-identifying face images
TLDR
A new privacy-enabling algorithm is presented, named k-Same, that guarantees face recognition software cannot reliably recognize deidentified faces, even though many facial details are preserved.
Efficient Privacy-Preserving Face Recognition
TLDR
A privacy-preserving face recognition scheme that substantially improves over previous work in terms of communication-and computation efficiency and has a substantially smaller online communication complexity.
Robust distance measures for face-recognition supporting revocable biometric tokens
  • T. Boult
  • Computer Science
    7th International Conference on Automatic Face and Gesture Recognition (FGR06)
  • 2006
This paper explores a form of robust distance measures for biometrics and presents experiments showing that, when applied per "class" they can dramatically improve the accuracy of face recognition.
Generating Cancelable Fingerprint Templates
TLDR
This paper demonstrates several methods to generate multiple cancelable identifiers from fingerprint images to overcome privacy concerns and concludes that feature-level cancelable biometric construction is practicable in large biometric deployments.
Blind Vision
Alice would like to detect faces in a collection of sensitive surveillance images she own. Bob has a face detection algorithm that he is willing to let Alice use, for a fee, as long as she learns
A Fuzzy Vault Scheme
TLDR
Fuzzy vaults have potential application to the problem of protecting data in a number of real-world, error-prone environments and also to biometric authentication systems, in which readings are inherently noisy as a result of the refractory nature of image capture and processing.
PICO: Privacy through Invertible Cryptographic Obscuration
  • T. Boult
  • Computer Science
    Computer Vision for Interactive and Intelligent Environment (CVIIE'05)
  • 2005
TLDR
This paper presents an example that demonstrates how using and adapting cryptographic ideas and combining them with intelligent video processing, technological approaches can provide for solutions addressing these critical trade-offs, potentially improving both security and privacy.
Fuzzy Extractors: How to Generate Strong Keys from Biometrics and Other Noisy Data
TLDR
This work proposes two primitives: a fuzzy extractor extracts nearly uniform randomness R from its biometric input; the extraction is error-tolerant in the sense that R will be the same even if the input changes, as long as it remains reasonably close to the original.
Capacity and Examples of Template-Protecting Biometric Authentication Systems
TLDR
A general algorithm is presented that meets the requirements and achieves Cs as well as Cid (the identification capacity) for privacy protecting biometric authentication systems.
...
1
2
3
4
...