IN-DEPTH

The article is devoted to understanding the evolution of SSML (Speech Synthesis Markup Language), the W3C standard for TTS behavior control. After SSML 1.0 became a W3C Recommendation in March 2004, public requests were collected to extend it for internationalization usages.
The W3C Voice Browser working group took this request seriously and organized three workshops in different regions of the world: Beijing - China (2005), Crete - Greece (2006) and Hyderabad - India (2007). In the meantime a new SSML 1.1 subgroup was created, with a wide selection of experts from different countries and started to produce results.
The article will describe this process and will give the first insights into the new features to be covered by SSML 1.1.

>> Go to the article.

STANDARDS CORNER

• SISR Published as W3C Proposed Recommendation
  February 5, 2007
  The SISR (Semantic Interpretation for Speech Recognition) specification has been published as a Proposed Recommendation. This is the stage before becoming a W3C Recommendation. SISR has demonstrated its implementability with the publication of an Implementation Report (Loquendo is one of the three implementers). Moreover SISR addressed all the public comments received.
  When the SISR 1.0 becomes a W3C Recommendation, hopefully next month, another important standard for the speech market will be put in place. The speech grammars will be completely standardized, both the SRGS syntax, already a W3C Recommendation since March 2004, and now the semantic interpretation, which is the part of a speech grammar devoted to producing the recognition results.
  SISR allows the use of the well-known ECMA-script language, also known as JavaScript, to generate and validate recognition results. Moreover the results can be cleanly represented in XML format, if the ASR interface requires it (e.g. MRCP protocol).
  
  
• SSML 1.1 First Documents have been Published
  January 10, 2007
  The work done in the W3C Voice Browser Working Group on a new 1.1 version of SSML has been recently yielding the first results. On December 19, the SSML 1.1 Requirements document was published and on January 10, the very first SSML 1.1 Working Draft.
  The editors of SSML 1.1 are Dan Burnett (Nuance) and Zhi Wei Shuang (IBM). The goal of this activity is to allow SSML to be used in a large variety of languages, which means extending it to better meet the major needs of these languages. The more complex issues and the ones that will require experts outside the authors will be deferred to a future major version of SSML.
  
  
• Summary and Minutes of the SSML Workshop III in India
  January 13-14, 2007
  The W3C Voice Browser Working Group has organized a third Workshop on "Internationalizing the Speech Synthesis Markup Language" in Hyderabad (India), in order to solicit additional suggestions for increasing the use of SSML in other languages.
  The workshop was kindly hosted by Bhrigus Sotfware and the International Institute of Information Technology (IIIT) in Hyderabad, India.
  W3C recently released a public summary of the workshop and the minutes of the discussions. More than 15 experts were present from India, Sri Lanka, Pakistan, Japan, Italy, US, France, and the UK.
  
  

Loquendo at SpeechTEK WEST

R&D COLLABORATIONS

• Loquendo Voice Technologies in Services for NOmadic Workers (SNOW) - Project results
  In October 2006 the two year FP6 project "SNOW", aimed at providing services for nomadic workers in harsh industrial environments, was succesfully concluded.
  SNOW addressed the production and exploitation of simple or complex multimodal e-manuals. It fitted the requirements of aeronautic maintenance, a demanding case where mobile workers rely on procedures and often face tight time constraints in which to achieve high quality results.
  The consortium was composed of 7 partners: EADS Common Research Centre-France, SAP AG, Siemens Business Services - C Lab, Fraunhofer Institute, Loquendo S.p.A., Advanced Computer Vision GmbH, and the University of Technology in Graz.
  >> More
  
  
• OPPORTUNITIES for R&D COLLABORATIONS under EC's FP7
  On December 22nd, 2006 the European Commission opened a new Call for Proposals for submitting Research and Technological Development projects to be funded within the theme of Information and Communication Technologies (ICT). This is the first call within the 7th framework program for the specific "Cooperation" program launched by the EU. The call has a total indicative budget of more than 1000 million Euros to be assigned to the most promising and innovative projects submitted by May 8, 2007 (see the European IC homepage).
  Loquendo is interested in exploring new, stimulating and innovative research ideas involving voice technologies in any application field, together with its partners, customers, academic institutions and other interested parties, to be presented for the upcoming call. The strategic objectives of call 1 relevant to research in voice technologies and applications are:
  
  • Challenge 2 - Cognitive systems, interaction, and robotics
  • Challenge 4 - Digital libraries and content
  • Challenge 6 - ICT for mobility, environmental sustainability and energy
  • Challenge 7 - ICT for independent living and inclusion
  Loquendo, with its long tradition in advanced research activities and strong commitment in creating innovation in the voice market, has an outstanding history of successful R&D projects that have been approved and funded by the European Commission in recent years. Currently Loquendo is involved in nine FP6 projects (C@R, COMPANIONS, COVER, DIVINES, HIWIRE, HOPS, I-WAY, LUNA, SHARE).
  To learn more about Loquendo's ongoing and previous European research projects please visit our international cooperation page.
  
  
• DECALOG workshop - Call for Papers
  DECALOG, the workshop on the Semantics and Pragmatics of Dialogue, is being jointly organised by the LUNA (Spoken Language UNderstanding in multilinguAl communication systems) project, of which Loquendo is coordinator.
  DECALOG is the 2007 edition of the SEMDIAL series of workshops and will be held at the University of Trento in Rovereto (Italy), May 30th - June 1st. It is being held in conjunction with the inaugural workshop of the Language, Interaction and Computation Lab from the Center for Mind/Brain Sciences (CiMeC), which will be on May 29.
  The deadline for submitting papers for review is 18 February 2007. For more information, click here.
  

LOQUENDO Products

• New Release of Loquendo Embedded ASR
  January 23, 2007
  Loquendo announced the release of the new embedded speech recognition 7.0.6, which includes many key new features. Loquendo ASR Embedded gives integrators the freedom to create user-friendly embedded applications that can be as complex as they want in terms of vocabulary size and interaction flexibility.
  Loquendo Embedded ASR is the right choice for PDAs, Tablet PCs, Automotive Telematics and any embedded application that needs to hear your voice. Loquendo ASR 7.0.6 supports Pocket PC 2003, Windows Mobile 5.0 for Pocket PC and Windows CE 5.0 for ARMv4i and x86 microprocessor based systems, as well as the software developing environment Visual Studio 2005.
  >> More
  
  

Clients & Partners

• The 'A Impecável' Department Store Launches Automatic Dialler Incorporating Loquendo Speech Technology Integrated by CTI Dealer
  February 12, 2007
  Loquendo and CTI Dealer, the computer-telephony integration company based in São Paulo, Brazil, announced the incorporation of Loquendo TTS into an automatic dialler for the 'A Impecável' chain of department stores in Rio de Janeiro.
  The automatic dialler makes use of Loquendo's high quality text to speech technology to personalise its messages. Automated collection calls are made to the store's charge card customers to remind them that the payment date has passed, and Loquendo TTS customises the message by reading out the name of the card holder.
  >> More
  
  
• Loquendo Technology Integrated into Vida-Software's Natural Interaction Technology
  January 17, 2007
  Loquendo and Vida Software, the Spanish multinational that has pioneered Natural Interaction technology, announced the integration of Loquendo ASR and TTS into Vida-Software's Natural Interaction Technology for mobile phones.
  Natural Interaction is a technology that allows users to talk and listen to a mobile service or application whilst simultaneously viewing or writing on the screen. The use of Loquendo's highly expressive Text-to-Speech and robust and accurate Speech Recognition combined with the ability to use multiple interaction tools concurrently, along with other Natural Interaction biometric technologies , result in an enhanced and satisfying user experience, both for existing applications and services as well as for new ones. Information is accessed by touch, speech, sight and sound, and this multi-sensory interface greatly simplifies phone operation.
  >> More
  
  

EVENTS

• SpeechTEK WEST 2007,
  Hilton San Francisco, BOOTH #314 Grand Ballroom, Hilton & Towers - San Francisco, California, 21-23 February 2007.
  
  
• ISS World 2007,
  Crowne Plaza Dubai, Dubai, UAE, 25-28 February 2007.
  
  
• ACE 2007,
  The Marriott Marble Arch, London - UK, 12-14 March 2007.
  
  
• Vocalexpo 2007,
  CNIT, La Défense, Paris, 13 March 2007.
  
  
• CeBIT 2007,
  Hannover, Germany, Hall 13, STAND C10, 15-21 March 2007.
  
  
• Servicios de Atención al Ciudadano,
  Hotel Melia, Avenida de America, Madrid, Spain, 27-28 March 2007.
  
  
• NAVIGATION EUROPE 2007,
  Marriott Hotel, Amsterdam, the Netherlands, 6-7 June 2007.
  
  
• SpeechTEK 2007,
  Marriott Marquis, New York, BOOTH #509, 20-23 August 2007.