January 24th, 2008

Loquendo, provider of speech technology worldwide, launches the latest version of its speech recognition engine - Loquendo ASR 7.5 - with many significant improvements.

The new features of Loquendo ASR 7.5 include:

• improved performance with complex grammars and large vocabularies
• support for 16 kHz acoustic models
• support for multilingual grammars
• differentiated timeouts for utterances of fixed format and length, e.g. credit card numbers
• improved confidence values enhancing performance in noisy environments
• support for Word Spotting

Each of these features has resulted in a more rapid, more robust and more efficient performance of Loquendo's speech recognition engine: even when using particularly complex grammars or grammars with large vocabularies, Loquendo ASR performs extremely effectively thanks to a reduction in the number of grammatical nodes and a redesigning of compiled recognition objects (ROs).

Loquendo ASR 7.5 allows the usage of 16 kHz acoustic models, e.g. for use with microphones on PCs, PDAs, navigation devices, etc.

Multilingual grammars are now possible because Loquendo ASR 7.5 supports the importing of foreign words into grammars: foreign words are transcribed according to the foreign language's phonetic rules and are then mapped onto the main language's acoustic units by means of Loquendo's patented Phonetic Mapping technique, already successfully deployed in Loquendo TTS and winner of the AVIOS-SpeechTEK award.

Differentiated timeouts are now supported - useful for recognizing, for example, a fixed number of digits (e.g. credit cards) - allowing extended support for the VoiceXML 2.0 'incompletetimeout' and 'completetimeout' properties. This speeds up results completion, returning recognition results faster as soon as a match in one of the active grammars has been found.

Loquendo ASR 7.5 displays dramatically enhanced recognition performance when noise is present, e.g. outdoors, in-car and warehouse, thanks to two new implementations: a noise compensation feature that results in improved confidence values in noisy environments, and furthermore, all supported languages have been re-trained with additional material recorded in the presence of background noise and via mobile.

Loquendo ASR is now available in 19 languages, with several recent additions - Turkish, Canadian French, Valencian and Galician.

The full list of supported languages for LASR 7.5 is: US English, UK English, Argentinian, Chilean, Mexican and Castilian Spanish, Catalan, Valencian, Galician, French, Canadian French, Portuguese, Italian, Dutch, Polish, Turkish, German, Greek and Swedish as well as Mixed Language Castilian Spanish and Catalan.

Like all Loquendo products, Loquendo ASR is available for both Windows and Linux operating systems, and in the embedded version for all major embedded operating systems.

Meet the Loquendo TTS family

About Loquendo - Vocal Technology and Services
With over 30 years experience in speech technology, Loquendo is at the forefront of the global speech market. Loquendo TTS, Loquendo ASR, Loquendo Speaker Verification and Loquendo VoxNauta are high-quality, high-performance technologies that guarantee systems integrators the best solutions whatever the application.

Loquendo is the only speech technology vendor that provides a complete product line from desktop PCs and telephony servers to navigation and handheld devices, guaranteeing the same portfolio of languages and the same core engine in these environments.

Many enterprises and carriers worldwide have recognized the Loquendo difference and our advanced technologies power millions of calls every day in the telecommunications and enterprise markets across the globe.

Loquendo is a Telecom Italia company headquartered in Turin, Italy, with offices in the U.S., Spain, Germany and France, and a worldwide network of partners. For more information about Loquendo, and to hear demonstrations of all Loquendo voices, visit www.loquendo.com.