Navigation Content

VoCon® 3200

Advanced Continuous Speech Recognition

 

VoCon® 3200, Nuance’s advanced speech recognition engine, is the benchmark for high-quality, speaker-independent, continuous speech recognition. The VoCon 3200 engine provides a broad range of unique features, including:

  • Highly accurate recognition of natural, conversational input in a broad range of languages
  • Ability to handle extremely large tasks such as destination entry for an entire country
  • Support for complex dynamic content, such as music titles, in over 25 languages

VoCon® 3200 State-of-the-Art Speech Recognition Engine

  • Noise robust front-end
    Tuned for a wide range of car environments, including mass-market saloons, SUVs, light commercial vehicles and convertibles.
  • Low CPU & RAM requirements
    Runs on processors and operating systems ranging from ARM 9 upwards. Its scalable architecture minimises RAM requirements.
  • Large vocabulary support
    Supports recognition of very large vocabularies for address entry, POI search and music selection by voice (300k+ word list).
  • One-shot destination entry
    Allows speaking a complete address, including house number, street name and city in a single utterance.
  • Natural Language Understanding (NLU)
    Recognises naturally spoken utterances. No restriction to pre-defined commands.
  • Music selection with partial title selection
    Recognises artist and title even if the user speaks only part of the name.
  • Support for multi-lingual input
    Recognises multi-lingual utterances
    for music and navigation applications which contain entries in multiple languages.
  • Industry's largest language portfolio Supports over 25 languages, including North American, Western European, Eastern European and Asian languages.

VoCon® 3200 Engine

Exceptional recognition accuracy is tuned for noisy automotive environments to deliver a superior user experience and high acceptance rates across a wide variety of in-vehicle systems and applications. Continuous improvements of the speech algorithms ensure highest accuracy standards even for large and complex tasks.

VoCon® 3200 Embedded Development System

Adds speech recognition functionality to any application. This includes VoCon® 3200 speech recognition engine, a robust set of development tools, guides and sample code that allow developers to build a high-quality speech-enabled application with optimum speed and efficiency.

VoCon® Toolkit

A set of utilities, meta functions and ready-made grammars accelerate development for the most common automotive use cases, including simple voice commands, phone dialling, music search and navigation destination entry.