ATOM Spoken Dialogue SDK for Embedded Systems

Contents

The ATOM Spoken Dialogue SDK Embedded includes unique powerful tools to support the development of advanced spoken dialogue applications.

• Grammar Compiler
  The grammar compiler translates grammars written in one of the supported grammar standards to an efficient binary representation. Please see Developers page for supported standards.

• Dialogue Script Compiler
  The dialogue script compiler translates dialogue scripts written in ECMAScript into an efficient binary representation. The output of the dialogue script compiler defines how the ATOM Dialogue Runtime Engine acts and reacts. This output can be redistributed or put in ROM.

• ATOM Dialogue Runtime Engine
  A dialogue runtime engine for Windows XP SP2 or higher is included in the SDK. Additional runtime engines for different target platforms can be obtained separately. Please see the Developers page for other target platforms.

• Database Conversion Tool
  The database conversion tool lets you convert SQL databases between different formats. Currently supported formats are Sqlite and MS Access.

• ATOM Grammar and Dialogue Tool
  The ATOM Grammar & Dialogue Tool is an Integrated Development Environment (IDE) for developing spoken language applications. In addition, Wizard-of-Oz based data collection, transcription and usability evaluation is supported.

• ATOM Data Collection Tool
  The ATOM Data Collection Tool is a simple tool to support non-Wizard-of-Oz based data collection for grammar development. It allows the developer to get an initial data set quickly, without having to invest a large amount of time and resources.

• ATOM Lexicon Tool
  The ATOM Lexicon Tool lets a developer interactively determine phonetic transcriptions of words. The phonetic transcriptions can be used to tune a speech recognizer or to add new words to the lexicon of the speech recognizer.

• Example Source Code
  The included example source code illustrates how dialogue scripts can take advantage of the features of the ATOM platform. In addition, C++ example source code shows how the ATOM Dialogue Runtime Engine can be integrated with third party software components.

Advantages

The ATOM Spoken Dialogue SDKs are complete software development kits for quick development of an advanced speech interface. The SDK contains all necessary development tools, environment and the ATOM dialogue engine.

Fast Development

• Comprehensive Infrastructure:
  Quick integration with most speech recognition and Text-to-Speech software, databases, and application programs.

• Grammar Tool:
  Automatic grammar error tester and comfortable editor make the development of speech recognition grammar faster than ever.

• Widespread use of industry standards:
  Guarantees a short learning curve for developers by supporting most established industry standards, including W3C SRGS (WML, ABNF), JSGF, W3C SISR, ECMAScript, RDFS.

• Open Platform:
  The openness of the platform enables its use both in research and development environments, while simultaneously allowing or a quick transition from prototype to market.

Target OS

Compatible Operation Systems

The ATOM Dialogue Runtime Engine is the part of the application that is responsible for natural language understanding, dialogue management and natural language generation. It is the ATOM Dialogue Runtime Engine that needs to be integrated with your application in order to make it understand spoken language.

The ATOM Dialogue Runtime Engine is available for several different operating systems and hardware platforms. Currently, the following platforms are supported:

Included in SDKs:

• Windows XP SP2 or higher
• Windows Vista
• Windows 7

The ATOM Dialogue Runtime Engine for other target OS is offered separately. Versions for the following OS are available:

• Windows Mobile 5.0, 6.0 or higher
• Windows CE
• Windows Automotive
• Linux x86, Kernel 2.6.10 or higher
• Linux ARM, Kernel 2.6.10 or higher

The ATOM Dialogue Runtime Engine can be easily ported to different operating systems or hardware platforms. If your targeted platform is not listed here, please contact us to explore the possibilities of a customized port.

Requirements

Development machine

• Operation System:
  Windows XP SP 2, Windows Vista or Windows 7

• Memory:
  At least 512 MB

• Hard disk:
  At least 500 MB free

• Additional Hardware:
  Microphone and sound card for speech input
  Loudspeaker for speech output

• Software:
  Microsoft Visual Studio 2005, 2008 or 2010
  Cygwin (free download)

Target machine

• Processor:
  x86 compatible or ARM compatible
  400 MHz or higher

• Memory:
  16 MByte free or higher

Customization

Customization of the ATOM Dialogue Runtime Engine

The ATOM Dialogue Runtime Engine is highly customizable to suit your needs. We are happy to discuss your project specific requirements with you. Below are some examples showing how the ATOM Dialogue Runtime Engine can be customized.

• Runtime Engine Customization

  We offer services to customize our highly modular ATOM Dialogue Runtime Engine to suit the specific needs of your application. Customization may include the addition or removal of functionality, as well as integration with your applications.

• Speech Recognition and Text-To-Speech Engine Integration

  The ATOM Dialogue Runtime Engine allows the easy integration of third-party speech recognition and text-to-speech engines.

• Port to different Operating Systems and Hardware Platforms

  To meet the requirements of your specific application, we offer custom ports of the ATOM Dialogue Runtime Engine to different operating systems and hardware platforms.

• Complete Spoken Language Application Design Development

  In addition to the above services, we also offer the complete design and development of spoken language applications based on our ATOM platform.