Virtual Facial Expression Dataset UIBVFED

Facial expression classification requires large amounts of data to reflect the diversity of conditions in the real world. Public databases support research tasks providing researchers an appropriate work framework. However, often these databases do not focus on artistic creation. We developed an innovative facial expression dataset that can help both artists and researchers in the field of affective computing. This dataset can be managed interactively by an intuitive and easy to use software application. The dataset is composed of 640 facial images from 20 virtual characters each creating 32 facial expressions. The avatars represent 10 men and 10 women, aged between 20 and 80, from different ethnicities. Expressions are classified by the six universal expressions according to Gary Faigin classification.

La classificació de les expressions facials requereix de grans quantitats de dades per reflectir la diversitat de condicions que demanen les aplicacions reals. Les bases de dades públiques acceleren les tasques de recerca proporcionant als investigadors un marc de treball adequat. Tradicionalment aquestes bases de dades no estan enfocades a la creació artística. Aquí presentem una base de dades formada per 640 imatges facials corresponents a 20 personatges virtuals executant 32 expressions cada un. Els avatars representen 10 homes i 10 dones de diferents edats i ètnies racials. Les expressions estan classificades segons les 6 expressions universals i per cada una d’elles segons la classificació que en fa Gary Faigin.

Link to the virtual app.

Link to download the data set

Link to the article

Contact: uibvfed@uib.cat

Citation. Oliver MM, Amengual Alcover E (2020) UIBVFED: Virtual facial expression dataset. PLoS ONE 15(4): e0231266. https://doi.org/10.1371/journal.pone.0231266