STV Reference: IRs 1488, 1489
Automated Speech Recognition Technology Limited by Bandwidth Requirement:
Over the past decade Automated Speech Recognition (ASR) systems installed on personal computers and workstations have achieved a high degree of accuracy. ASR systems have been integrated with web browsers to create voice-enabled web browsers. For example ASR systems can be used to have the computer ""speak"" the content of a web page, or to enable the user navigate the web by speaking the links that the user wants to follow. However, these functions are limited because the bandwidth required to transfer binary sound files over the Internet is extremely large, the web browser can only speak to the content of the web page, the audio content is generated from web pages primarily designed for visual interpretation, and the visual interface is removed from the user’s experience.
Voice Activated Web Pages for Hand Held Devices:
This technology includes: (1) system for processing voice request; and (2) computer usable medium storing voice request processing program. Specifically, a HTML request specifying a web page is received, and the presence of rule-based grammar statement in the received request is identified. When a connection associated with identified grammar is specified in data structure, the request is processed.
Applications:
For processing voice request for presenting voice activated web page in client device such as cellular telephone, personal digital assistant (PDA) and tablet personal computer (PC).
Advantages:
This technology enables processing voice request effectively and rapidly, thus addressing the need to independently create an audio component of a web page that does not demand a large amount of transmission bandwidth and exists in conjunction with the visual component of a web page.
Patent Status: Patent Pending (
US 20050143975)
Licensing Status: Available for Licensing and Sponsored Research Support