Face Annotation

A new report, "Face Annotation at the Macro-scale and the Micro-scale: Tools, Techniques, and Applications in Forensic Identification", is now available from the NIJ through the National Criminal Justice Reference Service.

Title:
Face Annotation at the Macro-scale and the Micro-scale: Tools, Techniques, and Applications in Forensic Identification (pdf, 8 pages)

Authors:
Patrick J. Flynn, University of Notre Dame; Anil K. Jain, Michigan State University

Abstract:

The authors developed a robust facial mark detection method that provides both location and scale information of facial marks. The report shows the utility of facial marks in face image retrieval tasks using a large-scale face image database. The proposed facial mark-based matching scheme can be combined with commercial off-the-shelf face matchers to improve the overall matching accuracy. The improvements of face-matching accuracy are shown both with a general face image database and an identical-twin face image database. The authors proposed a conditional filtering scheme to reduce the retrieval time on a large-scale face image database. This study on symbolic query-based face image retrieval shows the great utility of facial marks when a query image is not available. Although the study shows a number of cases where facial marks can be used for face image retrieval tasks, the instability of facial marks is still a major obstacle to using facial marks in practice.

Download the report here

 
< Prev   Next >






Crime Scene Revisited

Faces of the victims recovered from the scene of a genocide.

Read more...