Speech recognition and text to speech provider Loquendo (News - Alert), has announced the launch of its latest version of Loquendo ASR 7.5.

 

The speech recognition engine now boasts a number of improvements and features to make it possible to more quickly and efficiently create and utilize advanced, complex speech.

 

Available for both Windows and Linux operating systems, and in the embedded version for all major embedded operating systems, ASR 7.5 makes it much easier to deal with complex grammars and large vocabularies, and thanks to reduced grammatical nodes, and a redesign of recognition objects, enhances its overall performance. 

 

The latest version also includes support for multilingual grammars including the importing of foreign words into grammars thanks to Loquendo's phonetic mapping technique which transcribes foreign words according to that language's phonetic rules and then maps it into the main language's acoustic units.

 

To enhance recognition performance, the latest version includes a noise compensation feature and has re-trained all supported languages to include background noise and mobile use.

 

Recognition results are also returned at a much faster rate with the latest edition as it now supports differentiated timeouts for recognizing something like a fixed number of digits (e.g. credit cards) and allowing support VoiceXML (News - Alert) 2.0 'incompletetimeout' and 'completetimeout' properties.

 

According to a company news release, Loquendo ASR 7.5 can be used with microphones on PCs, PDAs, navigation devices and others as it allows the usage of 16 kHz acoustic models.

 

 

For more, be sure to visit the Speech Recognition and Text to Speech channel on TMCnet.

Stefania Viscusi is an established writer and avid reader. To see more of her articles, please visit Stefania Viscusi’s columnist page.