Replies to post #13379 on e.Digital Corporation (fka EDIG)

[cksla]

June 24, 2002 Speech Recognition Market Talking to Dollars in 2001 and Beyond

Despite an economic climate that was particularly damaging to new technologies and the financial difficulties of Lernout and Hauspie, one of the leading providers of the technology, the US speech recognition market continued to grow in 2001, according to In-Stat/MDR. The high-tech market research firm reports that, while not the watershed year of 2000, the market, as a whole, continued to embrace speech recognition products in 2001 as a means of lowering costs, providing greater customer service, and launching new, innovative, products. During the year, market leaders such as Speechworks, IBM, and Philips Speech increased penetration, while the previous market leader, Nuance, experienced a 24% decrease in revenues.

We are just on the cusp of a speech recognition revolution. Carriers are only now adopting the products, the technology has just reached the point of combustion, and standards such as Voice eXtensible Markup Language (VXML) and Speech Application Language Tags (SALT) are poised to change the velocity of the market. Given these factors, speech recognition will find its way to the mainstream and very soon we will be using it in our homes.

In-Stat/MDR believes that the following will contribute to the continued growth in adoption and pervasiveness of speech recognition:


The use of standards such as VXML and SALT, which will open up vast new markets and opportunities.

Continuing increases in processor speeds that will allow for greater speech recognition vocabulary size and accuracy.

The importance of measuring Return on Investment (ROI) in a tight economy, and voice recognition’s ability to adequately demonstrate it.

Significant advances in development tools and the resulting increase in production speed and lower costs.

Carrier demand for churn reduction and greater customer revenue.

An increase in the number of wireless subscribers.

Increased regulation of the use of wireless handsets while operating a car.

[cksla]

cksla In reply to: mike.w who wrote msg# 1800 Date: 5/8/2001 11:40:36 AM
Post # of 13385

TI tries to secure the wireless future
By Richard Shim
Special to CNET News.com
May 7, 2001, 5:25 p.m. PT
A new wireless technology will allow consumers to make secure purchases on next-generation cell phones and wireless handhelds, Texas Instruments said Monday.

The new security technology is a software library for TI's blueprint for wireless devices that use its digital signal processors. The blueprint is known as OMAP, or the Open Multimedia Applications Protocol.

"At TI, security is a religion. And to keep themselves in the No. 1 spot in the DSP (digital signal processor) market, the company is already addressing new services for the next generation of wireless phones," said Will Strauss, president of market researcher Forward Concepts.

TI's efforts will allow consumers to perform activities over devices with high-speed Internet access, such as buying products, downloading and streaming content, performing bank transactions and interacting with corporate networks.

The software library consists of programs from securities companies such as SafeNet, NTRU, WhiteCell, SnapShield and AuthenTec. The applications provide capabilities such as memory protection, advanced public and private key encryption, virus screening, firewall protection and fingerprint identification.

Security has already been tagged as a key feature for wireless devices.

"In Europe and Asia where (next-generation) phones are already being used, customers are already complaining about being spammed and compromised," Strauss said. "TI is just getting ahead of the game."

Multimedia has also been identified as a key feature for future wireless devices. But TI Chief Executive Tom Engibous recently warned that manufacturers should remember to keep features simple.

Handset manufacturers such as Nokia, Ericsson, Sony, Sendo and HTC have selected OMAP as the framework for their next-generation phones.

[cksla]

TI showcases its 2.5G/3G applications developers
By John Walko
CommsDesign.com
(12/20/01, 01:10:29 PM EDT)

Texas Instruments has brought together 25 of the companies that use its OMAP platform and processors to develop applications for 2.5 and 3G phones, wireless PDAs and mobile Internet appliances at a showcase event held at its European headquarters near Nice, France.

The companies, all members of the OMAP Developer Network, demonstrated applications for multimedia, audio, games, speech, location based services, messaging, security and mobile commerce.

The OMAP platform is TI's family of wireless modem and application processors for mobile phones and mobile Internet appliances. The processors support all the popular high-level operating systems and programming languages, including Symbian's EPOC, Microsoft Windows CE, Java, Palm and Linux, and the platform is already used by companies such as Nokia, Ericsson, Sony, Sendo, LG and Acer in their wireless devices. The latest OMAP application processor, the 1510, is based on TI's 320C55xx digital signal processor core and an enhanced ARM processor.

At the event, TI highlighted the $100m investment fund to help developers accelerate the creation of applications for third generation mobile networks, which it launched at the GSM World Congress in Nice in February. However, TI would not say how much of the $100m earmarked for the investment fund has been spent or allocated, or how many software developers and technology partners it has invested in as part of the project. To qualify, small, entrepreneurial groups have to deploy the OMAP platform to develop applications.

Companies demonstrating their developments at Nice included:

*BitFlash, the Ottawa, Canada based mobile software developer, which revealed at the event that it is collaborating with TI to deliver the first application of mobile scalable vector graphic (SVG) technology on the OMAP platform. The BitFlash Reflexis technology, once ported to the OMAP, will allow handset makers to render and transmit scalable interactive graphics and animations such as mobile postcards, as well as create a high quality graphical user interface.

Donna Ronayne, vice president of business development at BitFlash, said her company's technology already delivers sophisticated graphical content that performs adequately. “But by combining this Reflexis technology with the OMAP processors will greatly improve the quality of mobile applications and content, offering high performance SVG rendering on mobile devices while minimizing power consumption”.

*PacketVideo, the San Diego, California headquartered specialist in wireless multimedia, demonstrated the Studio application development environment that allows developers to quickly assemble a host of streaming multimedia applications for next generation wireless devices based on the OMAP platform and PacketVideo's PVPlatform.

*Pace Soft Silicon, which is developing what it claims will be the first commercially available multimedia terminal for synchronized voice and video at low bit rates. The company, which has offices in London and Dublin and a development centre in Pune, India, concentrates on silicon resident software IP and integration services to speed OEM product development and reduce risk.

Paul Dyer, vice president of business development at Pace Soft Silicon, said, “our multimedia terminal solution is architected for maximum configuration, flexibility and re-use by OEMs building very complex products. The handling of real-time streamed data such as integrated audio, voice, video and graphics is a computationally intensive task, traditionally best executed on high performance DSP devices.

“Our architecture is optimised to run on the OMAP DSP, thereby incurring minimal processing overhead on the main application code running on the ARM core. Also, a broad range of multimedia functions can be dynamically configured and managed from the ARM application via our application-programming interface”.

*NTRU, an 18 month old company based in Burlington, MA., which is developing a high speed encryption, decryption and authentication system for the OMAP platform. NTRU says the security software helps eliminate the traditional trade-offs between price, performance and usability in mobile systems .The security can be processed simultaneously with the application (multi-tasking), enabling operations to run up to 125 times faster than any competing solution. The speed efficiency also reduces the drain on power consumption, memory and processing bandwidth.

*Eyematic, the Los Angeles, California based group developing a synthetic video platform for wireless applications. The development platform incorporates authoring tools that combine patented real-time facial tracking software with synthetic templates to make the visual creation process simple and inexpensive; and a delivery platform which integrates lightweight 2D and 3D animation formats with synthetic character encoding that is said to dramatically lower mobile infrastructure requirements.

Typical applications include two-way rich media enhanced SMS, picture mail, 2D &3D games, jokes, horoscopes and delivery of news to mobiles. Eyematic is backed by companies such as Deutsche Telekom, Omron and TRW and has already developed systems used by partners that includes Qualcomm, NTT DoCoMo, Sun Microsystems, Verizon Wireless, SK Telecom and Nokia.

*Advanced Recognition Technologies (ART), the Tel Aviv, Israel based specialist in embedded speech and handwriting recognition technology for wireless devices, whose human interface software is already used in over 19million handsets made by companies such as LG, Mitsubishi, Panasonic and Samsung.

*NeuVoice, a spin-out company from Neural Systems Ltd that was established to exploit research from the Centre for Neural and Adaptive Systems at the University of Plymouth in the UK, also demonstrated voice recognition technologies for wireless devices based on the OMAP platform. The company currently offers a development kit for the Symbian OS, with another for Windows CE to be launched soon.

NeuVoices's recognition engine is based on biological models of hearing and speech production, and the resulting algorithms code to an extremely small memory and processing footprint. The system does not require any special DSP or floating point processor.

http://www.csdmag.com/story/OEG20011220S0040