Product category:
Embedded Software and Operating Systems
News Release from: Aculab | Subject: CWR 3.0
Edited by the Electronicstalk Editorial
Team on 07 October 2002
Speaker verification software runs on
DSP card
Speech processing specialist Aculab is entering the biometrics arena with the launch of its own speaker verification and identification (SVI) software at SpeechTek 2002.
Speech processing specialist Aculab is entering the biometrics arena with the launch of its own speaker verification and identification (SVI) software at SpeechTek 2002 Designed for use in telephony applications, the new software forms a component part of Aculab's new connected word recognition (CWR) 3.0 package, which is available under a cost free licence when used with the company's Prosody digital signal processor (DSP) resource card
This article was originally published on Electronicstalk on 28 Feb 2002 at 8.00am (UK)
Related stories
IP gateway card is easy to listen to
Aculab has launched a 30-channel IP telephony gateway PCI card.
DSP-based speech synthesis speaks five languages
Aculab has unveiled the latest version of its highly acclaimed free text-to-speech (TTS) software, which now features five languages, and runs on its Prosody DSP resource cards.
Aimed at systems integrators and developers, the new SVI software uses statistical data to confirm a caller's identity by their unique voiceprint and provides an extremely cost effective, convenient and secure way to access personal information over the phone.
Because Aculab's SVI is built on existing speech recognition technology, the two can run simultaneously.
Ultimately, this means that applications can verify the caller based on their voice and for an even more secure verification, a piece of personal knowledge can be included in a recognition process.
Further reading
DSP-based speech processing aids teleconferencing
A revolutionary teleconferencing system that puts blind and partially sighted people in touch with each other has been developed for the RNIB by Chelston Call Systems.
DSP provides natural language speech recognition
A natural language speech recognition system at Singapore Computer Systems' JuzCall third-party call centre uses DSP resources from Aculab combined with speech recognition software from Nuance.
Using biometrics - or physical characteristics - of the human voice to verify that individuals are indeed 'who they say they are', Aculab's latest software offers high-density SVI that can run in parallel with its range of DSP resources.
These include record, playback, echo cancellation and dual tone multi-frequency (DTMF) detection to produce high quality versatile applications with low 'per channel' costs.
Working in conjunction with other speech processing technologies such as text-to-speech (TTS) and automatic speech recognition (ASR) further enhances self-service applications.
Not only can they process a caller's request and retrieve the right information, they can also verify the caller's ID.
In a typical application, the new software can provide high verification accuracy using only a single, short utterance.
Up to 120 channels can be supported by a 1.9GHz processor, widely used in new PCs.
Aculab's sales and marketing director, Chris Gravett, believes that the new SVI technology will be widely adopted by the financial services sector and government as well as by companies that need to give employees secure access to intranets, extranets and corporate applications.
"In the telecomms arena, it can be used to reduce fraud associated with calling cards", he says.
"For enterprises, the real advantage of SVI is the security issue that it overcomes.
It presents a huge cost saving for companies who have to spend money on sourcing lost PIN numbers and passwords".
• Aculab: contact details and other news
• Email this article to a colleague
• Register for the free Electronicstalk email newsletter
• Electronicstalk Home Page
Search the Pro-Talk network of sites
