Researchers develop way to identify first-person filmmakers from the camera wobble

Two researchers at the Hebrew University of Jerusalem have developed a way to identify first-person filmmakers based on the wobble of their cameras, according to a report by The Verge.

Professor Shmuel Peleg and Yedid Hoshen recently published their results in a paper titled “Egocentric Video Biometrics”, where it details how you can distinguish one person from another by analyzing an individual’s gait, or the unique pattern of their body movements.

The researchers tracked the “optical flow” of objects and surfaces, frame by frame, to successfully determine the identities of their volunteers based on only four seconds of first-person footage.

“This is like a fingerprint,” says Peleg. “In order to find the person you have to have their fingerprint beforehand. But we can compare two people and say whether two videos were shot by the same person or not.”

Peleg decided publish the findings to “make people think twice before uploading egocentric footage to the web”, pointing out that just because the person’s face isn’t shown on camera it doesn’t mean they are actually anonymous.

The researchers also said that law enforcement agencies might eventually be able to link first-person footage to video footage captured by CCTV.

Although the researchers have not yet tested this form of recognition, Peleg said when you view a person on a surveillance camera “you can see the way they move and the way they move their head”.

Based on this principle, it should be “possible” to connect the two, particularly in high-surveillance countries like the UK where there is one CCTV camera for every 11 people.

So far, the research is only based on head-mounted cameras. However, the researchers believe that the same methods could be used for body-mounted devices, which is especially notable in light of President Obama’s recent pledge to purchase 50,000 body cameras for US law enforcement.

Article Topics

behavioral biometrics  |  biometrics  |  cameras  |  gait recognition  |  identity verification  |  Israel  |  research and development  |  video