S.P.I.D.™
Speaker Identification and DetectionVoice Authentication
Relevant Media
Content Downloads
Contact Us
Have questions? Need more information?
The primary driver for Bell in implementing a voice verification system was to fight identity theft and make privacy protection more convenient for Bell customers across five service segments (wireline, wireless, high-speed Internet, digital TV and VoIP).
Monitor Communications and Capture Suspected Targets with Advanced Voice Mining and Speaker Identification
Intelligence and law enforcement agencies need the best available technologies to help them combat crime and continue to integrate advanced technologies as part of an overall effort to improve homeland security. Both sectors rely on extensive sources to track and capture targets, but to do so they must be able to effectively piece together relevant information from global communication networks and intelligence feeds to follow the right leads.
Nuance S.P.I.D. (Speaker Identification and Detection) is a comprehensive audio processing system that efficiently searches for a target’s voice within a large volume of intercepted calls, regardless of the conversation content or method of communication. S.P.I.D. directs law enforcement agencies and intelligence organizations to relevant calls containing a target’s voice. The system uses a unique biometric analysis that matches relevant audio segments of a call to a pre-enrolled voiceprint. S.P.I.D. provides an open architecture and flexible interfaces (via a Web service) that enable easy integration with any lawful interception system and monitoring center.
For more information, please contact salesvoicebiometrics@nuance.com.
Features & Benefits
Features
- Content, language and accent independent
- Single/multi-speaker call monitoring
- State-of-the-art accuracy
- Easy integration with any monitoring center
Benefits
- Enhanced operational efficiency
- Extended global coverage of intercepted calls
- Immediate, automatic speaker detection
- Increased privacy protection (focus only on relevant suspected calls)
Applications & Markets
Applications
Voice Mining
S.P.I.D. can be used to prioritize a series of intercepted calls based on the probability that a target speaker is a participant. Voice mining capabilities effectively monitor the communications of a suspected target and filter out all irrelevant calls, leaving only those calls with high confidence scores that are likely to contain the target's voice.
Speaker Identification
When an intelligence agency intercepts a call containing incriminating information, S.P.I.D. compares the voice of an unidentified speaker against a set of relevant voiceprints. A confidence score is given for each voiceprint, indicating the likelihood that the voice in the intercepted call matches one of the voiceprints, screening irrelevant voiceprints with low confidence scores.
Markets
- Intelligence Agencies
- Government and Law Enforcement
How It Works
How it Works
Intercepted calls are sent to S.P.I.D., which processes and scores them based on the probability that they contain the target's voice.
Enrollment
Enrollment requires a minimum of 30 seconds of net audio (target voice). When using stereo files, the system can extract the specified side containing a target speaker. Mono files can be edited manually to retrieve the target voice using advanced Web-based tools. When enough audio is accumulated, the voiceprint is created and stored in a secure directory.
Verification
Verification and identification are performed by comparing audio files to a single voiceprint (verification) or a list of voiceprints (identification), detecting the participating speaker(s).
