Technology

TECHNOLOGY PROFICIENCIES

SAA’s Speech Technology and Laboratory record in NextGen speech technology encompasses patents, products and a wide array of technical expertise. SAA voice recognition system, FLY-BY-VOICE™, employs AUTOMATION in developing novel applications of speech recognition technologies.

  • Complex and High-Noise Audio Environments: 
    Our FLY-BY-VOICE™ system recognizes speech in multi-party conversations and high noise environments such as flight environments, broadcast recordings and across a wide range of frequencies for radio communication channels.
  • Phonemic Indexing and Time Alignment:
    Our FLY-BY-VOICE™ system searches for and locates audio events of interest, making them retrievable.
  • Molecular Voice Recognition: 
    Our recognition engine integrates pitch and phonemic context to detect end of utterance rather than waiting for a set period of silence.
  • Speaker/Accent Independent Speech recognition: 
    Our technologies employ acoustic models and tools that are adaptable to any specialized population and desired dialect.
  • Lexicon Translation Systems: 
    Our FLY-BY-VOICE™ system accommodates interpretation of statistical translation for phrase-based or free-form spoken language input.
  • Fatigue Detection: 
    The advanced configurations of the FLY-BY-VOICE™ system incorporate methods and tools for classifying levels of alertness (sleep, low, awake, high alert) that human listeners can produce in speech.

FUNCTIONALITIES


Location Retrieval:

Function Navigation by Voice Command:
Data Entry:
Audio Mining and Time Alignment:
Molecular Speech Recognition:
Desktop speech recognition systems have been used either on a remote server or pre-loaded on a mobile device. Either approach forced users for environmental tradeoffs. FLY-BY-VOICE™ solution eliminates design sacrifices by using a unique molecular mode of speech recognition validated for robustness from high-stress environments to complex ARINC architectures.

SOFTWARE PLATFORM

SAA software platforms enable product engineers to embed speech recognition into their products and services. SAA FLY-BY-VOICE™ solutions include developer application programming interfaces and documentation, the speech recognition function and the Lexicon. FLY-BY-VOICE™ solution boasters 98% command accuracy performance, speaker independent speech recognition that scales from embedded small applications to large-scale architectures and integrated systems for use in industrial, consumer, and military products and systems.