Topic: Synthetically Template Transformation Network for Tattoo Retrieval
by Maciej Salwowski
D19/2.03a, January 18, 2024 (Thursday), 12.00 noon
Keywords — Biometrics, Tattoo Transformation, CNN, Tattoo Retrieval
Tattoos have been successfully used to assist law enforcement in the identification of offenders and victims. Due to various privacy issues in the acquisition of images containing tattoos, only a limited number of databases exist. This lack of databases has slowed down the development of new tattoo recovery methods. In our work, we present a novel deep neural network (DNN)-based architecture called Template Transformation Network (TTN) that uses synthetically generated tattoos to improve the identification performance of reference systems. TTN also exploits synthetic templates to reconstruct missing areas of the input tattoos and thus improve the intra-class and inter-class properties of the final embeddings. Experimental results on real data demonstrate the utility of synthetic template reconstruction for tattoo retrieval.