Voice Key technology is based on the uniqueness of the human voice, which carries individual information by virtue of the acoustic features of speech. They reflect both the peculiarities of the physiological structure of the speech apparatus and acquired habits. That's why recent development companies of biometric systems use voice to identify the speaker.
Application fields
- Information security systems of enterprise
- Telephone and WEB-services
- Voice passwords for access to Internet resources (e-banking, e-commerce, electronic subscription publications, etc.)
Tasks
- Organization of the regulated access of users on a given passphrase to enterprise resources, telephone and WEB-services
- Improving the quality of client service:
- Pattern generation – data on physiological or behavioral characteristics are transformed into a form that is available to computer technology and stored in biometric system’s memory
- Highlighting – unique indicators exploring by system are extracted from the newly presented identifier
- Comparison - compare information about a newly presented and the previously registered ID
- Solution - to make a conclusion on whether or may not coincide a newly presented and the previously registered ID
- Segmentation of customers according to their importance, setting priorities for groups or individual customers
- Getting on the information about the client by operator to start the conversation
The principle of operation consists of 4 main stages:
- Ability to identify a user, i.e. real confirmation of the authenticity of the person receiving access rights
- The security of systems is increased significantly and, at the same time, the process of user’s identification is simplified - the user does not have to remember, type and regularly change passwords in various systems
- Your safety - there is no possibility of losing or stolen the key. Your Voice - Your Key
- Due to the simplicity of authentication process, it can be triggered much more often than traditional systems allow (requesting user name and password)
- In all cases, except cracking, we can prove the authorship of an e-action, confirmed by voice identification/verification
- Complicate remote selection of identifying information
- High reliability and stability of the system
- Unable to show ID by a third party
Comparative characteristics of biometric identification and verification
| Fingerprint | Voice | Iris | Face | |
| Equal error level (EER) | 2–3.3% | 0.1 - 0.86% | 4.1-4.6% | 4.1% |
| Refusal to register | 4% | 2% | 7% | ~0% |
| Capture probability of false targets | 2.5% | 0.75% | 6% | 4% |
| Likelihood of missing targets | 0.1% | 0.75% | 0.001% | 10% |
| System cost | High | Low | Very High | High |
Distinctive features
- Low cost of implementation and ownership - does not require special expensive means of biometric information input
- Combination of communication and identification procedures
- High reliability verification, comparable to the verification with fingerprint
- Identification in telephone channel
- Independence from the national language or dialect
