Research Accomplishments of Latanya Sweeney, Ph.D.



Overview

Medical Informatics
      Scrub
      Datafly
      Genomic identifiability
      Patient-centered management

Database Security
      k-anonymity

Surveillance
      Selective-revelation
      Risk assessment server
      PrivaMix

Vision
      Face de-identification

Biometrics
      Contactless capture

Policy and Law
      Identifiability of de-identified data
      HIPAA assessments
      Privacy-preserving surveillance

Public Education
      Identity angel
      SSNwatch
      CameraWatch

Quantitative assessments

Vision: Face de-identification

[cite, cite, cite, cite, cite, cite, cite, cite]

Work done with my students, Ralph Gross 10 and Elaine Newton 11.

Problem Statement: Given video or a photograph, de-identify faces appearing in the video or photograph so that no matter how good face recognition software may become, it cannot reliably recognize the faces yet facial details remain in the image.

Description: The k-Same algorithm by Dr. Sweeney and students is a solution. It scientifically limits the ability of face recognition software to reliably recognize faces while maintaining facial details in the images. The algorithm determines similarity between faces based on a distance metric and creates new faces by averaging image components, which may be the original image pixels (k-Same-Pixel) or eigenvectors (k-Same-Eigen). Results are presented on a standard collection of real face images with varying k. We also show how ad hoc techniques (e.g., pixelation and additive noise) do not work. Later papers describe methods for producing photo realistic images in real-time using active appearance models and multi-factor models. Its privacy guarantee limits face recognition to do no better than guessing 1/k.

(a)

(b)

Scientific Influence and Impact: Dr. Sweeney and her students were first to demonstrate the importance of using provable privacy protection over ad hoc approaches, by showing how face recognition could re-identify faces distorted by masking, additive noise, and pixelation. They introduced the first formal model for protection. Others introduced alternatives and enhancements [Defaux et al.]. Senior recenty edited a book on the topic [cite]. Working with Gross, Cohn, de la Torre and Baker, they produced anonymized, photo realistic video of pain grimace in patients for NIH [cite].

Other Achievements: 12

  • Patent application filed.

  • The face de-identification paper of Dr. Sweeney and her students [cite] appeared in IEEE Transactions on Knowledge and Data Engineering (z-test: among the top 20% of computer science journals).

  • The face de-identification paper of Dr. Sweeney and her students [cite] appeared in an elite conference IEEE Conference on Biometrics (BTAS acceptance rate 10%).

  • The paper of Dr. Sweeney and her students [cite] follows the citation trend modeled by my most cited k-anonymity paper [cite] (Spearman rank correlation rs=1.00, statistically significant at 99.9% confidence level). Using this correlation, a 4 year estimate predicts a citation count, statistically significant at the 99.9th percentile, among those of Associate Professors in the School of Computer Science at Carnegie Mellon University.



Notes

10 Ralph Gross graduated with PhD and advanced to a technology start-up and a part-time faculty position at Carnegie Mellon University.

11 Elaine Newton advanced to a position at NIST.

12 See quantitative assessments for more details.

Previous | Next


Related links:


Fall 2009