You are here

Demo

Speech Input Plate

  • Provide the most fast and precise Speech Input Solution
  • One technical milestone of Speech Lab in Large Vocabulary Continuous Speech Recognition (LVCSR)


Mobile Checkin

This project proposes authentication system with high accuracy of information, which is a combination of voice, avatar, location, and other multi-modal. We have applied the latest voiceprint recognition and face recognition technology to mobile terminals, and with help of advanced positioning technology, we achieve high-precision movable terminal authentication.

The entire authentication system consists of three modules: a voiceprint recognition module, face recognition module and integrated authentication module.

The voiceprint recognition module consists of advanced voiceprint identification system, which is able to capture the features of speaker with short voice recordings, with high training speed and accuracy.

The face recognition module uses Face++ cloud computing platform, using the industry's most efficient recognition algorithm, which is stable and reliable.

In the Comprehensive validation module, the results from voiceprint recognition and face recognition module become input. And through the GPS positioning system, we strictly control the identity verification.

Innovative Features:

  • Safety and reliability: the system contains an innovative combination of voice, avatar, location, and other multi-modal information, and the authentication methods has been through large amounts of data adjustment.
  • Low cost: the entire system contains only a single server, except an ordinary mobile terminal, and based on the advanced algorithms, even an ordinary computer can do the job.
  • Ease of deployment: the server and the mobile terminal software can be installed simply and quickly.

The Qr code is as follows: