Original video frame Tampered video frame

What is “Video Authentication”?

Video authentication is a process that is used to ascertain the trustworthiness of a digital video. In other words, a video authentication system ensures the integrity of digital video, and verifies that the video taken into use has not been tampered.

Motivation

We motivate the need for authenticating a digital video by giving the following examples:

A video clip can be doctored to defame a person. On the other hand, criminals get away from being punished because the video showing their crime can’t be proved trustworthy in the court of law.

In surveillance systems, it is hard to reassure that the digital video produced as evidence is the one that is actually shot by the camera.

A journalist cannot prove that the video played by a news channel is trustworthy.

A video viewer who receives video through a communication channel cannot ensure that video being viewed is really the one that was transmitted.

So there is a compelling need for video, wherever it is and in whatever form it is, be made authenticable before use.

This research

In this research, we propose a scalable signature scheme for video authentication. The proposed video authentication scheme is sensitive to spatial and temporal tampering; and also robust to frame dropping. The core idea of our technique is to utilize three hierarchical levels of a video and to use cryptographic secret sharing to create what we call as a ‘secret frame’. We authenticate a given video by computing the secret frames based on randomly generated private key at three hierarchical levels i.e. key frame level, shot level, and video level. We first segment the video into shots. Then, for each shot we identify the key frames. Here key frames refer to the frames that cannot be inferred semantically from other key frames within the same video shot. At the key frame level, we compute the secret for each pair of key frames using secret sharing considering all non-key frames between the two key frames as shares. The secrets computed at this level and the key frames are treated as shares to compute the secret at the shot level. Finally, all shot secrets are used to compute a master secret that is considered as the media signature (which we will refer to as the ‘signature’ from now onwards) for the video. The proposed method is scalable to three level of three levels of signature computation and can be tuned to the need of the different scenarios. In our scheme, the size of the authenticating signature is equal to a video frame size and it is independent of the length of the video in time.

Application scenarios

Our method can be used in the following three different kinds of scenarios:

In the scenarios where video is streamed through a communication channel, due to the large size of video data, the streaming often suffers from congestion problem at the bottlenecks on the network. To overcome the network congestion problem, some data loss (e.g. loss of few video frames) is common. For instance, the video transcoder or the designated router intentionally drops frames to save bandwidth or to avoid buffer overflow.

The proposed method is also useful in video identification. Video identification refers to a process that recognizes the existence of a particular video clip in large set of video data. For example, in an advertisement monitoring scenario where a commercial company or an individual can automatically identify in real time whether or not a TV channel is playing their video advertisement for the stipulated time. A TV channel may cut few frames to earn more time and the money. Our method can detect this type of tampering.

Detection of object/region (such as human faces) tampering in a video is another application where our method can be used. The proposed method localizes the important spatial regions in a video and assigns higher weights to them in the authentication process.

People

Pradeep K. Atrey, Wei-Qi Yan, and Mohan S. Kankanhalli

School of Computing, National University of Singapore, Singapore

Contact person: Pradeep K. Atrey

Publications

P. K. Atrey, W.-Q. Yan and M. S. Kankanhalli. A scalable signature scheme for video authentication. Springer Int. J. of Multimedia Tools and Applications, 34(1): 107-135 (2007).

P. K. Atrey, W.-Q. Yan, E.-C. Chang and M. S. Kankanhalli. A hierarchical signature scheme for robust video authentication using secret sharing. The 10th International Multi-Media Modelling Conference (MMM'04), pp 330-337, January, 2004, Brisbane, Australia.

© 2006, Pradeep Kumar Atrey, Wei-Qi Yan, Mohan S. Kankanhalli.  All rights reserved.