Vocalis announces port of Advanced Speech Recognition to Dialogic DM3

Vocalis, the Cambridge-based speech recognition telephony company, is joining with the Dialogic Corporation, one of the world’s leading manufacturers of components for the telecommunications industry to port it’s Advanced Speech Recognition (ASR) facilities into Dialogic’s DM3 computer telephony (CT) architecture.

The incorporation of ASR, based on the Vocalis leading edge Vocalis SpeechWare technology, within DM3 means that speech recognition capabilities will be tightly integrated in users’ telecommunications systems, making application development easier and faster. Also, the fact that ASR is not ‘bolted on’ externally helps to increase reliability and enhance robustness of performance.

"At Vocalis we have used Dialogic boards for our own speech recognition services, this development extends a link that already existed between us," said Mike Williams, Development Director for Vocalis. "This new global dimension to our partnership will bring together two market leaders in their respective fields".

"It is important that the speech recognition technology incorporated within DM3 is well proven in telecommunications markets, and Vocalis SpeechWare clearly meets that requirement," said Dialogic’s Pascal Dognaux, Product Manager for speech products at Dialogic. "We also wanted a partner that had extensive R&D capabilities in speech recognition, which Vocalis does. ASR technologies today are very demanding in terms of processing power and memory, and we are glad Vocalis see DM3 as the right platform."

Users of the DM3 architecture can develop their own customised speech processing systems using standardised Application Programming Interfaces (APIs). These will be available on either SCO Unix or Windows NT. In addition, Vocalis can provide customers with powerful toolkits tailored for developing Vocalis SpeechWare applications, such as recognition applications. For further support, Vocalis supplies development tools including Transcriba, which breaks down speech into components suitable for recognition, and Scripta, which helps designers to create an effective structure for interactive voice response (IVR) systems. Users will also be able to use industry standard development toolkits for application generation.

The combination of Vocalis SpeechWare and DM3 allows operations, like call centres, to rapidly develop sophisticated voice-driven services. These services include call routing, which means callers can be directed automatically to departments or individuals, or flexible information playout, in which appropriates messages can be played to callers relating to their particular location.

Dialogic calls DM3 a mediastream resource architecture. Rather than a specific product, DM3 consists of a board that acts as a hardware platform to which extra resources can be added to make deliverable products. Daughter cards can be added to provide extra processing power, DSPs and memory. Firmware technologies like ASR are then downloaded onto the DM3 hardware to create the deliverable CT platform.