SOTERIA

Description This dataset was used to perform the experiments reported in the IJCB2024 paper : "A novel and responsible dataset for presentation attack detection on mobile devices". The dataset consists of face videos captured using two cameras (main and front) of nine different smartphones : Apple iPhone 12, Apple iPhone 6s, Xiaomi Redmi 6 Pro, Xiaomi Redmi 9A, Samsung Galaxy S9, GooglePixel 3, Samsung Galaxy S8, iPhone 7 Plus, and iPhone 12 Mini. The dataset contains : Bona-fide face videos: 8400 videos of bona-fide (real, non-attack) faces, with and without hygienic masks. In total, there are 70 identities (data subjects). Each video is 10 seconds long, where for the first 5 seconds the data subject was required to stay still and look at the camera, then for the last 5 seconds the subject was asked to turn their head from one side to the other (such that profile views could be captured). The videos were acquired under different lighting conditions, including normal office lighting, low lighting, and outdoor lateral lighting. The data subjects were consenting volunteers, who were required to be present during two recording sessions, which on average were separated by about three weeks. In each recording session, the volunteers were asked to record a video of their own face using the front (i.e., selfie) camera of each of the five smartphones mentioned earlier. The face data was additionally captured while the data subjects wore plain (not personalised) hygienic masks, to simulate the scenario where face recognition might need to be performed on a masked face (e.g., during a pandemic like COVID-19). Attacks: Print attacks: 4665 video recordings of the data subjects’ face photos printed on two types of paper, matte and glossy, and two types of printers, laser and ink-jet. Replay mobile attacks : 11200 video recordings of bona-fide face videos that were replayed to the target smartphone’s camera. Different phones were paired, such that one of the pair was used to replay the bona-fide videos while the other (attacked) phone recorded the videos using its front camera. Replay TV attacks : 5600 video recordings of bona-fide face videos that were replayed on a TV screen. The recordings were performed under multiple conditions, including in normal and low lighting, as well as when the recording device was tilted. Projector attacks : 2800 video recordings of bona-fide face videos that were projected on a screen. The recordings were performed under multiple conditions, including in normal and low lighting, as well as when the recording device was tilted. Two projection screens were used, white and green.   Reference If you use this dataset, please cite the following publication: N. Ramoly, A. Komaty, V. K. Hahn, L. Younes, A. -M. Awal and S. Marcel, "A Novel and Responsible Dataset for Face Presentation Attack Detection on Mobile Devices," 2024 IEEE International Joint Conference on Biometrics (IJCB), Buffalo, NY, USA, 2024, pp. 1-9, doi: 10.1109/IJCB62174.2024.10744500.