Assessment of age estimation methods for forensic applications using non-occluded and synthetic occluded facial images

  1. Jeuland, Elouan Derenee 1
  2. Del Río Ferreras, Aitor 2
  3. Chaves, Deisy 2
  4. Fidalgo, Eduardo 2
  5. González-Castro, Víctor 2
  6. Alegre, Enrique 2
  1. 1 Ecole Nationale Supérieure des Mines de Saint Etienne
  2. 2 Universidad de León

    Universidad de León

    León, España


XLIII Jornadas de Automática: libro de actas: 7, 8 y 9 de septiembre de 2022, Logroño (La Rioja)
  1. Carlos Balaguer Bernaldo de Quirós (coord.)
  2. José Manuel Andújar Márquez (coord.)
  3. Ramon Costa Castelló (coord.)
  4. Carlos Ocampo Martínez (coord.)
  5. Jesús Fernández Lozano (coord.)
  6. Matilde Santos Peñas (coord.)
  7. José Enrique Simó Ten (coord.)
  8. Montserrat Gil Martínez (coord.)
  9. Jose Luis Calvo Rolle (coord.)
  10. Raúl Marín Prades (coord.)
  11. Eduardo Rocón de Lima (coord.)
  12. Elisabet Estévez Estévez (coord.)
  13. Pedro Jesús Cabrera Santana (coord.)
  14. David Muñoz de la Peña Sequedo (coord.)
  15. José Luis Guzmán Sánchez (coord.)
  16. José Luis Pitarch Pérez (coord.)
  17. Oscar Reinoso García (coord.)
  18. Oscar Déniz Suárez (coord.)
  19. Emilio Jiménez Macías (coord.)
  20. Vanesa Loureiro Vázquez (coord.)

Publisher: Servizo de Publicacións ; Universidade da Coruña

ISBN: 978-84-9749-841-8

Year of publication: 2022

Pages: 972-979

Congress: Jornadas de Automática (43. 2022. Logroño)

Type: Conference paper


Age estimation is a valuable forensic tool for criminal investigators since it helps to identify minors or possible offenders in Child Sexual Exploitation Materials (CSEM). Nowadays, Deep Learning methods are considered state-of-the-art for general age estimation. However, they have low performance in predicting the age of minors and older adults because of the few examples of these age groups in the existing datasets. Moreover, facial occlusion is used by offenders in certain CSEM, trying to hide the identity of the victims, which may also affect the performance of age estimators. In this work, we assess the performance of six deep-learning-based age estimators on non-occluded and occluded facial images. We selected FG-Net and APPA-REAL datasets to evaluate the models under non-occluded conditions. To assess the models under occluded conditions, we created synthetically occluded versions of the non-occluded datasets by drawing eye and mouth black masks to simulate the conditions observed in some CSEM images. Experimental results showed that the evaluated age estimators are affected more by eye occlusion than by mouth occlusion. Also, facial occlusion affects more the accuracy of the age estimation of minors and the elderly compared to other age groups. We expect that this study could become an initial benchmark for age estimation under non-occluded and occluded conditions, especially for forensic applications like victim profiling on CSEM where age estimation is essential.