IN-DEPTH

This issue's main article will introduce the readers to a detailed description of Automatic Speech Recognition in Loquendo ASR product. The goal of the article is to describe the main achievements and advances made in this field by Loquendo researchers, developments that have impacted strongly on speech recognition performance.

The nature of this contribution is technical and may require some knowledge of ASR development techniques.

>> Go to the article.

STANDARDS CORNER

• Voice Browser and Multimodal meetings in Oslo
  September 18-22, 2006
  The W3C Multimodal Interaction Working Group met in Oslo on September 18-19 and then jointly with the W3C Voice Browser Working Group on September 20. The W3C Voice Browser Working Group then continued with its meeting on September 21-22. The public summaries will be released in a short time on the public mailing lists.
  
  
• VoiceXML 2.1 Published Again Last Call Working Draft
  September 15, 2006
  Following the publication of this specification as Candidate Recommendation, substantial comments were received, and in accordance with the W3C Process, the Director returned the Voice Browser Working Group for further work. This work resulted in significant changes to the specification (see Section 6 on the use of <foreach> element), which led to a return as Last Call Working Draft.
  Public comments and reviews are encouraged on section 6 specifically, send them to the public Voice Browser mailing list www-voice@w3.org. The Last Call period ends on October 6th, 2006.
  
  
• New Working Draft of MRCPv2 Released by IETF
  September 15, 2006
  IETF announced that a new Internet-Draft is available from Speech Services Control Working Group (SPEECHSC) for "Media Resource Control Protocol Version 2 (MRCPv2)". The authors are D. Burnett and S. Shanmugham.
  The MRCPv2 protocol allows client hosts to control media service resources such as speech synthesizers, recognizers, verifiers and identifiers residing in servers on the network.
  MRCPv2 has reached a sufficient level of maturity to be a good candidate to be considered as a Proposed Standard. To read the new draft: http://www.ietf.org/internet-drafts/draft-ietf-speechsc-mrcpv2-11.txt. For more details on IETF SPEECHSC group, click here.
  
  
• Common Sense Suggestions for Developing Multimodal User Interfaces
  September 11, 2006
  The Multimodal Interaction working group published a W3C Note to describe the accumulated experience of several years of developing multimodal applications. It provides a collection of common sense advice for developers of multimodal user interfaces.
  
  

R&D COLLABORATIONS

• Loquendo Voice Technologies to Support Advanced eSafety Multimodal Applications for Road Transport
  Loquendo is providing speech technology for Cover, a European Union-funded project to make road travel safer, more environmentally friendly and more efficient.
  The COVER project will exploit (www.ist-cover.eu) many existing technologies to achieve a safer and enhanced travelling experience across Europe's road networks, involving cooperation between leading-edge European companies. The project will integrate technologies from many diverse fields, including Loquendo ASR and Loquendo TTS, in-car and infrastructure sensor data, context aware and multimodal interfaces, intelligent agents and semantic technologies.
  By improving communication channels between vehicles and the existing road network infrastructure, far more in-depth information will be passed on to the driver, such as fully up-to-date info on traffic conditions, recent road works and weather conditions. Further, by employing and sharing 'context knowledge', vehicles will have far more sophisticated recognition systems, directing drivers' attention to road hazard signs, sharp bends or accident hotspots, allowing appropriate action to be taken well in advance.
  Loquendo speech technology will provide a user friendly interface by which the driver can communicate with the numerous and varied technologies that will be integrated by the Cover project.
  >> More
  
  
• Kick-off meeting of LUNA project
  Loquendo hosted the kick off meeting of the LUNA (spoken Language UNderstanding in multilinguAl communication systems) project, co-funded by the European Commission under the 6th FP. Representatives from the following entities attended the meeting: University of Trento (IT), RWTH Aachen (DE), University of Avignon (FR), France Telecom (FR), CSI-Piemonte (IT), Polish-Japanese Institute of Information Technology (PL) and Polish Academy of Sciences (PL).
  The LUNA project addresses the problem of real-time understanding of spontaneous speech in the context of advanced telecom services, and aims at the creation of a robust natural Spoken Language Understanding toolkit for multilingual dialogue services, able to carry on human-computer communication with a good degree of user satisfaction.
  Further details about the project will be available in upcoming newsletters.
  
  
• Loquendo at the autumn Eurocities Knowledge Society Forum event organized by Torino City Council
  "The HOPS project - spreading out services through the information system of vocal recognition" will be presented by Franco Carcillo, manager of the ICT project, City of Torino and Joan Batlle, manager of the International Project, City of Barcelona.
  Within the 'eServices for eCitizen' workshop, Loquendo will share its expertise gained during the HOPS project with a speech on: "What can be done by voice today (commercial solutions and opportunities which can be deployed today) for municipalities and future capabilities of a new platform based on the HOPS findings on semantic enrichment of data, natural language processing and dialog management".
  To find out more about the Turin event, click here, for general informations about Eurocities (formerly Telecity), click here.
  
  
• Results of HIWIRE project at IST event in Helsinki
  Find out more about Loquendo R&D collaborations at the IST 06 event in Helsinki (21-23 November 2006), where we will show the results of the HIWIRE project.
  For further information visit the HIWIRE website: www.hiwire.org.
  
  

LOQUENDO Products

• Loquendo Presents Krzysztof, the Brand New Male Polish Voice
  September 27, 2006
  Loquendo is pleased to present Krzysztof, the new male Polish voice, as part of Loquendo's widely acclaimed text to speech technology. With the voice of Krzysztof, Loquendo's second Polish voice, the company extends its reach into the rapidly growing Central European markets, enabling the company to support its partners even more effectively in a diversifying European community.
  Krzysztof is the latest of our voices to be enriched with 'expressive cues', Loquendo's innovative and award-winning tool for achieving truly natural sounding speech. Loquendo is the very first speech company to offer an 'emotional' component, a major advance in speech technology giving Loquendo's voices truly lifelike expression and a far wider emotional range.
  To hear Krzysztof's natural sounding speech for yourself, click here.
  To hear Zosia for yourself, click here.
  >> More
  
  
• Loquendo ASR With New Tunable Parameters, Partial Results Recognition and Improved American English in Challenging Environments
  August 9, 2006
  Loquendo announced a set of new ASR features that enable integrators to design user friendly speech applications, which are as complex as they want them to be in terms of vocabulary size and interaction flexibility.
  Another feature of Loquendo ASR is the availability of Partial Results during the recognition process, capable of recognising a caller's utterances immediately, word by word, rather than needing to analyse the complete phrase. In this way, any incorrect formulations can be identified straight away, and the caller doesn't waste time completing futile input interactions.
  Furthermore, for American English, the recognition performance has been significantly improved, particularly in challenging environments.
  >> More
  
  
• Loquendo TTS Multimedia Package and Voice Creator Available
  August 8, 2006
  Loquendo announced the availability of Loquendo TTS in four convenient packages: Loquendo TTS Multimedia, Loquendo TTS Multimedia Compact, Loquendo TTS Telephony and Loquendo TTS Embedded. Four packages thoughtfully conceived to cater for the many and varied applications of speech technology, all running on the same core engine and all guaranteeing the same wide range of natural sounding voices in 18 languages.
  Loquendo offers two useful tools for creating your own text to speech audio files: Loquendo TTS Director and the new Loquendo TTS Voice Experience, both of which have been refined and expanded.
  >> More
  
  

Clients & Partners

• Loquendo's TTS Integrated into Ontira's IVR
  October 3, 2006
  Loquendo and Ontira, leading provider of Automated Traveler Information Systems (ATIS) technology, announced the integration of Loquendo's award-winning text to speech into Ontira's Interactive Voice Response solutions.
  Loquendo's TTS has been chosen for the exceptional emotional range of its voices and the high quality of its speech synthesis, and with Ontira's ATIS will enable travelers to access travel information and plan their journeys 24/7.
  By integrating Loquendo technology, Ontira will significantly reduce the costs associated with creating draft IVR systems by exploiting all the advantages of TTS, only requiring costly and time consuming recording once the IVR design is finalized at the closing stage. In addition, TTS will be invaluable for "on the fly" emergency bulletins, saving the time and money of having to make manual recordings.
  >> More
  
  
• Loquendo TTS and Loquendo ASR Integrated into COM plan + service 'Robot 5' Voice Portal
  September 19, 2006
  Loquendo and COM plan + service, leader in telecommunications systems, are pleased to announce the integration of Loquendo TTS and Loquendo ASR into the COM plan + service existing platform, 'robot 5'.
  Robot 5 is an easy-to-use development platform for building speech driven solutions, and is quick, intuitive, scaleable and manufacturer-independent. Robot 5 unites Loquendo's state-of-the-art technology with telecommunication hardware in a single application environment, and provides interfaces to information storage media such as databases and the internet.
  >> More
  
  
• Loquendo and the Universita' di Trento Announce a Collaboration to Support Education and Research on Speech Technologies
  September 6, 2006
  Loquendo and the Department of Information and Communication Technology (DIT), University of Trento - Italy, announced a long term collaboration aiming at supporting the Speech and Language Research laboratories for master and PhD students. This collaboration is devoted to give in-depth knowledge of the state of the art of TTS - Text To Speech, Automatic Speech Recognition - ASR, and spoken dialog applications, such as VoiceXML applications. The Speech Laboratory is already active and uses the most recent version of Loquendo's speech technologies and platforms.
  Loquendo and DIT will strictly collaborate on research topics to enhance speech technologies in advanced speech application. As an example, Loquendo and DIT will jointly address the real-time understanding of spontaneous speech in the context of advanced telecom services within the LUNA project.
  >> More
  
  

EVENTS

• Voice Days 2006,
  Bonn, Alter Bundestag, 18-19 October 2006.
  
  
• Expo Call Center Madrid 2006,
  Madrid, Feria de Madrid, Palacio Norte, 7-8 November 2006.
  
  
• Expo Call Center Estoril 2006,
  Estoril, Portugal, Estoril Congress Center, 28-29 November 2006.
  
  
• ISS World 2006,
  Hilton Crystal City, Washington D.C. - USA, 4-6 December 2006.
  
  
• SpeechTEK WEST 2007,
  Hilton San Francisco, BOOTH #314 Grand Ballroom, Hilton & Towers - San Francisco, California, 21-23 February 2007.