VoiceGrid™ RT

A sophisticated distributed solution for real-time speaker recognition in phone and other communication channels, from mobile devices and other sources. Integrates up to 10,000 target speakers into a voice database, which can be scaled up to millions of processed records per day.

VoiceGrid RT
Product Library:

Description

Having direct connection to communication channels this solution can identify a speaker on the 30th second of the dialog. We will choose components and build on the platform VoiceGrid RT any solution on the identification in the communication channels in real time.

Highlights

  • Simultaneous recording and voice biometric search
  • Recording and voice biometric search can be performed simultaneously so it helps to identify the speaker before the conversation completion.
  • Speech patterns examination
  • Integration with regional and national databases enables a specialist to start detailed examination of speech patterns in a few seconds after the procedures of monitoring and registration of suspicious phone conversations.
  • Data storage
  • All the audio-data is registered in the database, having template capacity of ten million recorded phone calls.
  • Sharing of data
  • The system allows to transfer all the records into the central database for the subsequent application at the other regions

Features

System mobility. A portable solution can be assembled on the basis of the VoiceGrid RT system and it can be connected to communication channels in any environment. In conjunction with portable cellular wiretapping equipment of the VoiceGrid RT system opens up new possibilities of operating voice search afield.

Quick results. The first results are received in a minute after the beginning of the telephone conversation. The accuracy is increasing due to the high volume of speech data accumulated. If the conversation continues for a long time, the result of a higher degree of reliability will be obtained before the conversation completion.

High reliability of the system. The first results are received in a minute after the beginning of the telephone conversation. The accuracy is increasing due to the high volume of speech data accumulated. If the conversation continues for a long time, the result of a higher degree of reliability will be obtained before the conversation completion.

Algorithms accuracy. High reliability of the results is obtained by application of three biometric algorithms. Alongside with the standard algorithms proprietary feature extraction and matching algorithms of Speech Technology Center are applied.

Simplicity of use. High degree of automation and simplicity of use allow training the customer’s staff in the process of solving the tasks by the system.

Complex solution. Speech Technology Center provides complex solutions for investigation specialists, as it combines multichannel phone recording and nation-wide voice biometric search.

Scalability

The system can be deployed in three modifications:

  • portable solution — 4-16 channels
  • standard solution — up to 60 channels
  • solutions on request — up to 1000 channels