Tohoku University. Research Profiles


Internet video streams have identities like fingerprints



According to recent studies, each video that is streamed online has its own unique characteristic. The pattern of data that are sent through the network can vary from one encoding method to another and the change from scene to scene, which is a unique characteristic of each video. This research aims to identify video content in the network without looking at the data itself.

Targeted Application(s)/Industry

This research would be especially useful in preventing illegal distribution of television program such as illegal channel without violating user privacy. Another application would be to prevent confidential online meeting from leaking outside local enterprise network. In addition, it is also possible to investigate the circulation of a video content using method introduced by this work.


Graduate School of Information Sciences

KATO Nei , Professor
Doctor of Engineering


Related Information

Hiroki Nishiyama, Desmond Fomo, Zubair Md. Fadlullah, and Nei Kato, "Traffic Pattern Based Content Leakage Detection for Trusted Content Delivery Networks," IEEE Transactions on Parallel and Distributed Systems, DOI: 10.1109/TPDS.2013.32, To be published.