We’ve noticed a few more press releases than normal that mention the granting of voice processing patents. Most recently, Advanced Voice Recognition Systems (AVRS) announced that it has been granted its second significant patent in the area of speech recognition and transcription. The company says that its first patent, U.S. Patent #5,960,447, was granted in September 1999. It carried the title “Word Tagging and Editing System for Speech Recognition.” The second patent, which was disclosed this month, addresses speech recognition and transcription “among users having heterogeneous protocols.”
We are neither lawyers nor patent experts but given that many of the commercial transcription services use a combination of speech processing and live agent intervention, this second patent is potentially very broad. In what could be construed as a threat to step up enforcement. Walter Geldenhuys, AVRS CEO and President, said that he believes the second patent “will enhance licensing and marketing opportunities.”
In a separate, and unrelated development, speech analytics specialist Utopy has been granted U.S. Patent #7,487,094 titled “System and method of call classification with context modeling based on composite words.” In this case, a company spokesperson told us that it is definitely “not an offensive patent.” Instead they see it as formal recognition of the fact that, in contrast to other speech analytics vendors, who take either a phoneme-based approach or resort to speech-to-text conversion before carrying out pattern recognition, Utopy’s methods are, indeed, unique.
It’s really a mouthful, but company co-founder Yochai Konig characterizes Utopy’s approach as “phrase-based contextual speech recognition-cum-understanding.” It provides the foundation for understanding live or recorded utterances in a single pass with both accuracy and completeness of understanding. It is an important component in business intelligence solutions that seek to link contact center and IVR performance with key business objectives. Utopy markets its solutions as premises-based, hosted or on-demand solutions.
Skype for the iPhone will be available for download from the iTunes AppStore tomorrow (Tuesday March 31). In the mean time the Twittersphere has been atwitter with links to reviews from the telco luminaries at CNET and GigaOm with reviews that have been universally favorable. For those, like me, who found the most value in using Skype from the PC to make inexpensive international calls, word is that “it’s all that!” and much more. So it is that a wide variety of both enterprise- and service provider-oriented infrastructure vendors have accelerated introduction of solutions sets that integrate screen-supported handling or origination of voice calls, voicemail, chat, conferencing and related call processing. 
In a post on AppleInsider reporter Aidan Malley is 
Attendees at eComm 2009 witnessed the coming-out party of a new technology firm called RebelVox as VP of Technology Matt Ranney ran through the core capabilities of the firm’s suite of software. Growing out of the defense communications domain, RebelVox has patent-pending technologies that, when combined, provide users with the ability to toggle seamlessly between synchronous (real-time) and asynchronous (recorded or retrieved) communications. Its as if individuals engaged in a phone conversation or conference call could take Tivo like control of both the origination and playback of content. 
Based on very favorable acceptance of a single-number service that started in October 2006, AirFrance entered the second phase of its speech enabled customer care portal “powered by Nuance”. The initial set of services offered flight schedules, bookings (with intelligent transfer to agents) and real time flight information in French and UK English. In the second phase new contact centers have been set up in Europe, Asia, North and South America supporting calls from over 20 countries.
Embedded speech processing specialist SVOX clearly believes that speech recognition may be a big part of any turnaround that could occur in the automotive industry. Stating that it shows commitment to growth, the company has established a new 30-person office in Ulm, Germany. CEO Martin Reber says that the staff will focus on development of embedded speech recognition and “speech dialog” systems. He expects the size of the Ulm-based team to double in the coming years as it takes advantage of opportunities to collaborate with firms like Audi, Daimler and Porsche on R&D efforts for speech processing in future models.