Authenticating a user subvocalizing a displayed text

A method of authenticating a user of a computing device, the method comprising:

• displaying a first text to the user,
• acquiring, from a camera operatively connected to the computing device, a video sequence of the user subvocalizing at least part of the first text,
• magnifying, by video processing the acquired video sequence, motions of one or more body parts of the user, which motions are correlated with the subvocalized at least part of the first text, and
• deriving a representation of the user subvocalizing at least part of the first text as a representation of the magnified motions from the video processing of the acquired video sequence,
• deriving a user phrasing signature by identifying a sequence of durations of pauses in the derived representation of the magnified motions from the video processing of the acquired video sequence, and
• authenticating the user in response to determining that the user phrasing signature and a reference phrasing signature fulfil a similarity condition, wherein the reference phrasing signature is derived from a previously acquired video sequence from the camera or another camera of the user subvocalizing the at least part of the first text.

A computing device for authenticating a user of the computing device, the computing device comprising processing circuitry configured to:

• display a first text to the user,
• acquire, from a camera operatively connected to the computing device, a video sequence of the user subvocalizing at least part of the first text,
• magnify, by video processing the acquired video sequence, motions of one or more body parts of the user, which motions are correlated with the subvocalized at least part of the first text, and
• derive a representation of the user subvocalizing at least part of the first text as a representation of the magnified motions from the video processing of the acquired video sequence,
• derive a user phrasing signature from the derived representation of the magnified motions from the video processing of the acquired video sequence,
• derive a reference phrasing signature by identifying pauses in the first text, acquiring reference durations of the pauses from a previously acquired video sequence from the camera or another camera of the user subvocalizing the at least part of the first text, and deriving the reference phrasing signature as a sequence of the reference durations of the identified pauses, and
• authenticate the user in response to determining that the user phrasing signature and the reference phrasing signature fulfil a similarity condition.