Philips Releases Speech SDK 3.1

VIENNA, AUSTRIA - Royal Philips Electronics released Speech SDK 3.1, a customizable software development kit for creating professional applications which benefit from the features of the Philips speech recognition technology. The new version includes a VoiceXML engine with dictation recognition (speech-to-text) and supports dictation input via telephone and microphone. Integrators can also take advantage of the range of services which Philips Speech Processing offers to customize Speech SDK to their specialist requirements. With the Speech SDK developers can integrate the full capabilities of the Philips speech recognition engine directly into their application software, using any standard software development environment through the PSP C/C++ API interface. Speech recognition capabilities include: dictation recognition, command recognition, spelling recognition, correction functions, natural language understanding and user interface components. New features of Speech SDK 3.1. are: ·Philips Enhanced VoiceXML: A Voice XML engine that includes dictation functionality. Speech SDK comes with a Natural Dialog Engine, which can process standard VoiceXML files which are typically used in the world of speech recognition to model a dialog between man and machine. Additionally it allows to access the dictation recognition directly from the VoiceXML script. ·Audio recording and playback: Speech SDK not only provides the recognition engine but also enables audio data to be recorded and played back. ·Reevaluation: In addition to HSA-4 (High Speed - High Accuracy) Speech SDK provides a special recognition method called reevaluation. This feature checks the recognition accuracy of critical parts and automatically adapts to changing sound conditions, thus increasing the recognition rate. ·Verification recognition: This feature verifies the spoken sentence against the original sentence and can be used to train users to pronounce phrases properly. This is an important feature for interactive language learning. ·Performance analysis: Speech SDK 3.1 includes special performance analysis functions to measure and improve recognition accuracy. You can use the performance analysis functionality to check and guide special adaptations of the speech recognition engine to suit your needs. ·Language support: In addition to the 21 languages supported so far, Philips is adding the Hungarian Command & Control language package to Speech SDK 3.1. Philips Speech Processing offers a Speech SDK Development Partnership program which includes consultancy services, technology adaptations designed to meet specific requirements as well as the opportunity to develop products jointly.