Solutions Automotive
       
Join our mailing list

for in-car multimedia and navigations

Multimodal in-car natural speech interface

Automotive Solutions Package is a complete solution for voice operated car navigation and in-car multimedia system. Original Equipment Manufacturers can develop cost-effective prototypes of spoken dialogue interfaces for their applications. The selection of appropriate questions to the user can be tailored to database results, the state of the car navigation system, or any other external information source.

Not speech recognition but speech dialogue

Our system can be integrated with most major speech recognizers, Text-to-Speech software, databases and avatars, and works on embedded and/or server based multimodal systems.

Please see the demo movie as well.

Car Navigation

Functions

General Dialogue Functionality

  • Multimodal (with or without touch screen, gesture camera and other smart tactile input interfaces)
  • Natural, free-form speech conversation
  • Multiple languages
  • Multiple turns of conversation
  • Barge-in speech

Automotive Functionality

Voice control for:

  • In-car media player
  • Voice dialing
  • Restaurant and POI search
  • Air condition

Service Included

  • Integration with ASR & TTS
  • Integration with your board
  • Integration with your webcode
  • Integration with your database
  • Integration with avater

Contents

  • Grammars (Source)
    The ATOM CarNavi Solution contains grammar for spontaneous spoken language understanding in both Japanese and English. Expressions covered include entering of Points of Interests (POIs), making phone calls and controlling in-car entertainment systems.
  • Dialogue Script (Source)
    The included dialogue scripts determine the way the system interacts with the user. Among other functionality, the scripts determine whether to ask the user for more information when entering POIs or making a phone call.
  • Database (source)
    The databases contain more than 500 POIs located in the New York Metropolitan Area (for the English version) and the Kyoto area (for the Japanese version).
  • Multimodal Framework (Source)
    The ATOM Multimodal Framework is a software layer on top of the ATOM Dialogue Runtime Engine that allows the fusion of input from speech recognition and touch screen. It simplifies the creation of application with multimodal input. Written in C++.