An investigation of Face Matching and Retrieval in Cyber Forensics Applications
Keywords:
Photo Matching, COTS Face-Recognition Systems-FRSAbstract
A latent fingerprint on the counter. A drop of blood on the floor. Law enforcement has successfully used these forensic clues to catch criminals for decades. But consider a face image captured by a surveillance camera that needs to be matched against millions of mug shots across the country. With the rapid increase in the number of surveillance cameras and mobile devices with built-in cameras, the forensics world is changing, and the progress in face recognition is helping to lead the way. In fact, in 2009, an estimated 30 million surveillance cameras were deployed in the US, shooting 4 billion hours of footage a week.[1 ] However, although recent research advances have helped lay the foundations for realizing face-matching scenarios for utilizing this data, face recognition in the forensics arena still poses a number of challenges. This article highlights the challenges in applying face-recognition technology to forensics applications. We explain why forensic face recognition differs from typical portrait face recognition and why a human examiner is often needed to carefully interpret and verify the matching results. Furthermore, we address three specific research problems that pose challenges to commercial-off-the-shelf (COTS) face recognition systems (FRSs):
- robustness to facial aging,
- retrieval using facial scars and marks, and
- matching forensic (composite) sketches to
- face photograph databases.
Solutions to these three problems are necessary to accurately remove duplicates in various
government face databases, including mug shot, passport, and driver license photos (aging-invariant FRS); to search a large face database when only partial or low-quality face images are available (scar and mark matching); and to apprehend criminals when no photo of the suspect is available (sketch to- photo matching). Additionally, we discuss methods that can augment existing COTS face-recognition systems by improving the quality of a face image prior to submission.
References
- J. Vlahos, 'Surveillance Society: New High-Tech Cameras Are Watching You,' Popular Mechanics, 1 Oct. 2009; www.popularmechanics.com/technology/military/4236865.
- S.Z. Li and A.K. Jain, eds., Handbook of Face Recognition, 2nd ed., Springer, 2011.
- P. Grother, G. Quinn, and P.J. Phillips, 'Report on the Evaluation of 2D Still-Image Face Recognition Algorithms,' Nat’l Inst. of Standards and Technology interagency/internal report (NISTIR) 7709, 2010; www.nist.gov/customcf/get_pdf.cfm?pub_id=905968.
- R. Charette, 'Here’s Looking at You, and You, and You. . .,' blog, IEEE Spectrum, 25 July 2011; http://spectrum.ieee.org/riskfactor/computing/it/ heres-looking-at-you-and-you-and-you-.
- L. Ding et al., 'Computers Do Better than Experts Matching Faces in a Large Population,' Proc. IEEE Int’l Conf. Cognitive Informatics, IEEE Press, 2010, pp. 280-284.
- V. Blanz and T. Vetter, 'Face Recognition Based on Fitting a 3D Morphable Model,' IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 25, no. 9, 2003, pp. 1063-1074.
- U. Park and A.K. Jain, '3D Model-Based Face Recognition in Video,' Proc. 2nd Int’l Conf. Biometrics (ICB), LNCS 4642, Springer, 2007, pp. 1085-1094.
- U. Park, Y. Tong, and A.K. Jain, 'Age Invariant Face Recognition,' IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 32, no. 5, 2010, pp. 947-954.
- T. Bourlai, A. Ross, and A.K. Jain, 'Restoring Degraded Face Images: A Case Study in Matching Faxed, Printed and Scanned Photos,' IEEE Trans. Information Forensics and Security, vol. 6, no. 2, 2011, pp. 371-384.
- B. Klare, Z. Li, and A.K. Jain, 'Matching Forensic Sketches to Mugshot Photos,' IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 33, no. 3, 2011, pp. 639-646.
- U. Park and A.K. Jain, 'Face Matching and Retrieval Using Soft Biometrics,' IEEE Trans. Information Forensics and Security, vol. 5, no. 3, 2010, pp. 406-415.
- Z. Li, U. Park, and A.K. Jain, 'A Discriminative Model for Age Invariant Face Recognition,' IEEE Trans. Information Forensics and Security, vol. 6, no. 3, 2011, pp. 1028-1037.
- 'Los Angeles Officials Identify Video Assault Suspects,' New York Times, 7 Jan. 2011; www. nytimes.com/2011/01/08/us/08disabled.html.
- B. Klare and A. K. Jain, 'Heterogeneous Face Recognition using Kernel Prototype Similarities,' tech. report MSU-CSE-11-18, Michigan State Univ., Nov. 2011.
- H. Ling et al., 'Face Verification Across Age Progression Using Discriminative Methods,' IEEE Trans. Information Forensic and Security, vol. 5, no. 1, 2010, pp. 82-91.
- X. Wang and X. Tang, 'Face Photo-Sketch Synthesis and Recognition,' IEEE Trans. Pattern Analysis and Machine Intelligence, vol. 31, no. 11, 2009, pp. 1955-1967.
- H. Tuthill and G. George, Individualization: Principles and Procedures in Criminalistics, Lightning Powder Company, 2002.
- B. Klare and A.K. Jain, 'On a Taxonomy of Facial Features,' Proc. 4ith IEEE Int’l Conf. Biometrics: Theory, Applications, and Systems (BTAS), IEEE Press, 2010, pp. 1-8.
Downloads
Published
Issue
Section
License
Copyright (c) IJSRCSEIT
This work is licensed under a Creative Commons Attribution 4.0 International License.