Enabling greater safety and convenience, VoCon 3200 is ideally suited for navigation systems, telematic applications, and command and control of in-vehicle devices. VoCon 3200 has been designed to improve the human-machine interface, placing special emphasis on voice activated dialing (VAD) and voice destination entry (VDE). In addition to speaker-independent continuous digit and command recognition, it supports text enrollment of address books to allow speaker-independent name dialing with no training required. What's more, VoCon 3200 can speech enable navigation systems, supporting street and city name recognition, even of large countries like the United States or of European wide navigation systems. Other VoCon 3200-driven applications include car environment control, MP3 player control, PC and console games, warehousing applications, and more.

Designed to be modular and highly scalable, VoCon 3200 flexibly meets the varying requirements of different application scenarios and markets. Modules like speaker adaptation, natural language understanding (NLU), spelling post-processor, and online grammar compiler for dynamic vocabularies can optionally be deployed on the target platform.

A special focus has been set on high recognition accuracy. The VoCon 3200 engine provides a broad range of advanced capabilities that help you deploy embedded applications delivering new levels of speech recognition accuracy even in noisy environments like Automotive.

• Leverage the power of an industry-leading embedded speech recognition system that has been proven to deliver maximum accuracy—even in noisy application areas like Automotive.
• Take advantage of the modular, scalable VoCon 3200 architecture for speech-enabling a variety of applications ranging from simple dialing to voice destination entry from a database of hundreds of thousands of names.
• Easily port the VoCon 3200 code base to embedded and automotive platforms supporting a large number of processors and operating systems.
• Accelerate application time to market and maximize deployment quality by using the VoCon 3200 speech recognition engine in combination with Nuance's broad range of development tools, services, and consulting offerings.

VoCon 3200 ASR Development Tools complement typical graphical user interface (GUI) dialog tools. While GUI tools are effective for accelerating development of relatively straightforward speech User Interfaces (UIs) that involve a small vocabulary, clean speech, native speakers, etc., they simply lack the capacity to rapidly pinpoint and improve problematic states in the dialog.

The VoCon 3200 ASR Development Tools offering is divided into two separate, but integrated tool suites: the Grammar and Pronunciation Editing and Analysis Suite and the Recognition Analysis Suite. The Grammar and Pronunciation Editing and Analysis Suite is designed to speed up all phases in the speech User Interface (UI) design cycle—from rapidly creating simple grammars, through early testing of the speech recognition performance, to thorough analysis of the more difficult, complex states in the dialog. Using these tools will reduce costs during the VUI prototyping, testing, and tuning phases. The tools help to develop and test grammars, pronunciation dictionaries, and engine parameters quickly and easily even without a deep knowledge of the VoCon 3200 recognition engine characteristics and parameters.

The Recognition Analysis Suite is used to analyze and improve recognition performance and to collect speech data, identify problems, and remove the causes of possible recognition errors.