C-14: assured timestamps for drone videos

Zhipeng Tang, Fabien Delattre, Pia Bideau, Mark D. Corner, Erik G. Learned-Miller

Abstract: Inexpensive and highly capable unmanned aerial vehicles (aka drones) have enabled people to contribute high-quality videos at a global scale. However, a key challenge exists for accepting videos from untrusted sources: establishing when a particular video was taken. Once a video has been received or posted publicly, it is evident that the video was created before that time, but there are no current methods for establishing how long it was made before that time. We propose C-141, a system that assures the earliest timestamp, tb, of drone-made videos. C-14 provides a challenge to an untrusted drone requiring it to execute a sequence of motions, called a motion program, revealed only after tb. It then uses camera pose estimation techniques to verify the resulting video matches the challenge motion program, thus assuring the video was taken after tb. We demonstrate the system on manually crafted programs representing a large space of possible motion programs. We also propose and evaluate an example algorithm which generates motion programs based on a seed value released after tb. C-14 incorporates a number of compression and sampling techniques to reduce the computation required to verify videos. We can verify a 59-second video from an eight-motion, manual motion program, in 91 seconds of computation with a false positive rate of one in 1013 and no false negatives. We can also verify a 190-second video from an algorithmically derived, 4-motion program, in 158 seconds of computation with a false positive rate of one in one hundred thousand and no false negatives.