The Text Recognition Algorithm Independent Evaluation (TRAIT)

Published

December 15, 2017

Author(s)

Afzal A. Godil, Patrick J. Grother, Mei L. Ngan

Abstract

The report describes and presents the results for text detection and recognition (TRAIT) evaluation in support of forensic investigations of digital media. These im- ages are of interest to NIST’s partner law enforcement agencies that seek to employ text recognition in investigating this area of serious crime. Our first evaluation uses images seized in child exploitation investigations. The primary application is the iden- tification of previously known victims and suspects, as well as detection of new vic- tims and suspects. The presence of text, for example, on a wall poster or on an item of clothing, may allow a location to be identified and linked to prior cases. In total, 3 groups took part in this evaluation over three Phases. The evaluation results show that the initial performance of text recognition is low. However, from Phase 1 to Phase 3, the performance of text recognition algorithms has shown improvement. We hope this evaluation will stir more research in this field.

Citation

NIST Interagency/Internal Report (NISTIR) - 8199

Report Number

8199

NIST Pub Series

NIST Interagency/Internal Report (NISTIR)

Pub Type

NIST Pubs

Download Paper

DOI Link

Keywords

Text detection, text recognition, evaluation, competition, benchmarking

Citation

Godil, A. , Grother, P. and Ngan, M. (2017), The Text Recognition Algorithm Independent Evaluation (TRAIT), NIST Interagency/Internal Report (NISTIR), National Institute of Standards and Technology, Gaithersburg, MD, [online], https://doi.org/10.6028/NIST.IR.8199 (Accessed December 11, 2024)