Biometric data has been used around the world for identification for a number of years. Typical applications include verifying identity at immigration and customs counters, and fingerprints to unlock mobile devices. Despite its wide application, once the biometric scan data is hacked or stolen, the owner can’t change his or her fingerprints and has to find another identity security system.
With this in mind, a scholar at HKBU has invented a new technology that uses lip motion password recognition. This system uses a person’s lip motions to create a password. It then verifies a person’s identity by simultaneously matching the password content with the underlying behavioral characteristics of the lips’ movement. A user’s lip movement when speaking the password can’t be mimicked and the password can be changed at any time.
Science Professor Cheung Yiu-ming from HKBU led the research. He listed a number of advantages the new technique has over conventional security access control methods:
- Verification based on a combination of password content and lip motions result in access control that is doubly secure;
- Lip motions are resistant to mimicry due to their dynamic characteristics. A lip password can therefore be used singly for speaker verification, as it is able to detect and reject a correct password uttered by an imposter or the wrong password spoken by the user;
- A user can reset the lip password at any time to reinforce security;
- The acquisition and analysis of lip movements is less susceptible to distance and background noise when compared to traditional voice based authentication. It can also be used by a person with a speech impairment;
- As there is no language boundary, a person from any country can use this verification system.
Yiu-ming noted that the same password spoken by two people is different and a learning system can distinguish between them. The study employs a computational learning model that extracts the visual features of lip texture, shape and movement to characterise lip sequence. To train the models and determine the threshold of rejecting and accepting a spoken password, samples of lip sequences are collected and analysed.
The new technology has been patented and potential applications includes transactions at ATM machines, financial transaction authentication including electronic payment using mobile devices and credit card user passwords. It can also be applied to enhance access control systems currently used in entrances of private or company premises.
Lip passwords can also be used in combination with other biometrics to enhance the security level of systems. Lip passwords can for example be combined with facial recognition. This would solve the problem of facial recognition systems being spoofed with 3-D masks.