A dataset of face images, collected by the Biometrics Vision and Computing (BVC) group
The face dataset is not available for cloning
Due to the license agreement involved in acquring the face dataset, a formal request is first required before releasing the dataset.
A dataset of Nigerian face images - BVC-UNN-face data, was collected by the Biometrics Vision and Computing (BVC) group at the University of Nigeria in 2018. We are making the BVC-UNN-face data free and publicly available with the spirit of promoting the sharing of academic knowledge. The BVC group through the University of Nigeria retains ownership and copyright of the BVC-face dataset.
This work is a study of face recognition performance using a database of faces of Nigerian subjects drawn from 28 different ethnicities.
Face videos with both expression and head pose variations were collected in two major sessions between June and November, 2018. 534 subjects participated in the first session. In the second session, 151 subjects participated; 134 subjects participated in both sessions. The face videos were captured at distances within 0.5 meter to 1.2 meters. High-resolution frames of face expressions and face poses, were grabbed from the videos.
The facial expressions include the following eight (8) conditions: neutral, laughing, smiling, frowning, fearful, squinting, screaming, and surprised. The facial poses include frontal, elevated or lowered pitch, and right or left yaw.
The dataset is based on the publication:
O. Iloanusi, P. J. Flynn and P. Tinsley, "Similarities in African Ethnic Faces from the Biometric Recognition Viewpoint," 2022 IEEE/CVF Winter Conference on Applications of Computer Vision Workshops (WACVW), Waikoloa, HI, USA, 2022, pp. 419-428, doi: 10.1109/WACVW54805.2022.00048. https://ieeexplore.ieee.org/document/9707558
The BVC-face data is distributed via the University of Notre Dame.
A formal request is required to use the BVC-face data. In other words, a researcher requesting for the dataset is required to fill a formal agreement and return to us via his/her institutional email.
Request to use the face database can be sent to any of the following emails: