SULFA: Surrey University Library for Forensic Analysis

SULFA is an open library for benchmarking video forensic algorithms. It was designed mainly for camera identification and integrity verification, but will be further extended for other forensic purposes. SULFA contains original as well as forged video files, which are freely available at the University of Surrey's website. Currently, SULFA is composed of around 150 videos collected from three cameras: Canon SX220 (H.264), Nikon S3000 (MJPEG) and Fujifilm S2800HD (MJPEG). Each video is approximately 10 seconds long with resolution of 320x240 and 30 frames per second. All videos were shot after carefully considering both temporal and spatial video characteristics. In order to present life-like scenarios, various complex and simple scenes were shot with and without using a camera support (tripod). SULFA also includes some manipulated videos with cloning or copy-paste forgery. For each forged video, full information of the doctored regions is also provided.

Additional IFS-related resources are available on the IFS TC website.