TattTRN: Template Reconstruction Network for Tattoo Retrieval

TL;DR

This paper introduces TattTRN, a novel network that improves tattoo retrieval accuracy by reconstructing tattoo templates, using a large semi-synthetic dataset to address privacy issues and enhance discriminative features.

Contribution

The paper presents a new unsupervised generative approach for creating a large tattoo database and introduces TattTRN, a network that maps tattoo images to templates to improve retrieval performance.

Findings

Achieved up to 99% accuracy in tattoo retrieval tasks.

Created a semi-synthetic dataset of 28,550 images across 571 categories.

Demonstrated effectiveness on real tattoo databases.

Abstract

Tattoos have been used effectively as soft biometrics to assist law enforcement in the identification of offenders and victims, as they contain discriminative information, and are a useful indicator to locate members of a criminal gang or organisation. Due to various privacy issues in the acquisition of images containing tattoos, only a limited number of databases exists. This lack of databases has delayed the development of new methods to effectively retrieve a potential suspect's tattoo images from a candidate gallery. To mitigate this issue, in our work, we use an unsupervised generative approach to create a balanced database consisting of 28,550 semi-synthetic images with tattooed subjects from 571 tattoo categories. Further, we introduce a novel Tattoo Template Reconstruction Network (TattTRN), which learns to map the input tattoo sample to its respective tattoo template to enhance the distinguishing attributes of the final feature embedding. Experimental results with real data, i.e., WebTattoo and BIVTatt databases, demonstrate the soundness of the presented approach: an accuracy of up to 99% is achieved for checking at most the first 20 entries of the candidate list.