RT info:eu-repo/semantics/conferencePaper
T1 Impact of facial occlusions in age estimation algorithms for forensic applications
A1 Carofilis Vasco, Roberto Andrés
A1 Chaves, Deisy
A1 Martínez Mendoza, Alicia
A1 Fidalgo Fernández, Eduardo
A1 González Castro, Víctor
A1 Alegre Gutiérrez, Enrique
A2 Blanco FernándezYolanda
A2 Fernández VeigaManuel
A2 Fernández VilasAna
A2 FuentesJosé M. de
A2 Ingenieria de Sistemas y Automatica
K1 Cibernética
K1 Ingeniería de sistemas
K1 Age estimation
K1 Deep Learning
K1 Facial occlusion
K1 CSEM
K1 3304.05 Sistemas de Reconocimiento de Caracteres
K1 1207.03 Cibernética
AB [EN] Age estimation is a critical tool for identifying minors and potential offenders in Child Sexual Exploitation Materials (CSEM). However, deep learning methods, which are the current state-of-the-art for age estimation, have low accuracy in predicting the age of minors and older adults due to a lack of data in existing datasets. In addition, facial occlusion, used to hide victims’ identities in some CSEM, can further impact the accuracy of age estimators. This study evaluates the performance of six deep learningbased age estimators on facial images with and without occlusion. To simulate CSEM conditions, we used non-occluded datasets (i.e., FG-Net and APPA-REAL), and also synthetically occluded versions with black masks of these datasets. Results indicate that age estimators are more affected by eye occlusion than by mouth occlusion and facial occlusion affects minors and the elderly more than other age groups. This study could serve as a reference for the estimation of age in forensic applications such as victim profiling on CSEM.
PB Universidade de Vigo
SN 978-84-8158-970-2
LK https://hdl.handle.net/10612/20673
UL https://hdl.handle.net/10612/20673
NO Vasco, R. A. C., Sánchez, D. C., Mendoza, A. M., Fernández, E. F., Castro, V. G., & Gutiérrez, E. A. (2023). Impact of facial occlusions in age estimation algorithms for forensic applications. In Actas de las VIII Jornadas Nacionales de Investigación en Ciberseguridad: Vigo, 21 a 23 de junio de 2023 (pp. 497-498). Universidade de Vigo.
DS BULERIA. Repositorio Institucional de la Universidad de León
RD 29-jun-2024