Home | Contact Us | FAQ | Search & Site Map | Link to Us
Sign In | Join | Other 45 Sites in Network
Home
Discussion GroupsEnglish UsageBritish EnglishESL Teaching
Learnglish.com
Contact UsLink To UsSearch & Site Map

Discussion Groups / ESL Teaching / October 2004



Tip: Looking for answers? Try searching our database.

New Technology from Sensory allows English Lessons on Handheld Devices

Thread view: 
Enable EMail Alerts  Start New Thread
Thread rating: 
Sensory Marketing - 23 Sep 2004 23:44 GMT
Sensory Inc, the world leader in embedded speech technologies, is now
demonstrating its new patent-applied-for English language training
technology that is implemented on low cost chips. The new technology
is being introduced in Hong Kong at Sensory's "Designing a Voice User
Interface" seminar.

Chinese Market Opportunity
The Chinese government has mandated that all service employees will
learn English by the 2008 summer Olympics in Beijing. This has caused
a rapidly growing market to emerge for handheld language training
devices. Typically, these language trainers will use chips from
companies like Sensory or Scansoft to play samples of the English
phrases and to allow a Chinese speaker to compare his/her recorded
voice with that of the English speaker.  Sensory's new technology goes
one step beyond this to perform an analysis of the Chinese speaker's
phrases so the speaker can get immediate feedback on the
pronunciation. It's much like an English instructor on a chip!
According to Sensory's CEO, Todd Mozer "Sensory is already selling
millions of chips each year into this rapidly growing market, and
these chips are just used for playback of speech, not analysis or
feedback. Sensory's language training technology is exactly what our
customers want!"

How it Works
A phrase is recorded into a handheld device and the technology
analyzes the phrase's phonemic content, which is compared with that in
stored templates of the properly spoken phrase. Playback of the
recorded phrase can be sped up or slowed down, and an LED or LCD
display can provide feedback on the quality of the spoken phrase
during playback in order that the user can visually assess the
performance. Settings can be adjusted to switch between a "lenient"
instructor and a "tough" instructor, and training phrases can be
cycled through to find specific phrases or general areas to practice
(e.g. shopping, business, socializing, etc.).

The RSC-4x Family & Tools
Sensory has chosen to rollout and showcase its language training
technology on the RSC-4x Family. This family of integrated circuits
(IC's) has the ability to support a wide variety of speech
technologies on a general purpose 8-bit microcontroller core. The
RSC-4x Family is supported by a full suite of tools including,
assembler, C-Compiler, debugger, and voice coding tools (QT2SI
vocabulary generator, Quick Synthesis speech compressor, etc.).

Edgar Chau, Managing Director of product development house
Cyber-Workshop says "We are familiar with the RSC-4x Family of IC's,
and we see much demand from our customers for this new language
training technology. The Beta unit we received from Sensory worked
quite well and we are sure to have many designs for 2005".

Sensory plans to bring its language training technology to other
embedded platforms,  including the ARM line of microcontroller
products popular in cell-phones.

Availability & Pricing
Sensory has created a demo platform that showcases a product design
concept for a low cost language-trainer. Sensory is seeking strategic
partners to roll out English language training handheld products
around the world. The complete system is expected to retail for
$19.95-$49.95 depending on features.

About Sensory, Inc.
Sensory, Inc. is the world leader in embedded speech technologies.
Sensory's speech technologies include language training, speech
recognition, speaker verification, speech synthesis, and animated
speech, and are used in consumer electronics, cell phones, PDA's,
Internet appliances, interactive toys, automobiles, and other
applications where low cost and high quality is essential.  Sensory
offers a complete line of integrated circuit and embedded software
solutions to customers such as Hasbro, JVC, Kenwood, Matsushita,
Mattel, Mitsubishi, Sharper Image, Sony, Toshiba, Uniden, and many
others.  Sensory is a profitable privately held company founded in
1994. More information can be found at www.sensoryinc.com.
Peter Trompetter - 16 Oct 2004 17:00 GMT
Hi,
Great solution!  You should only make it different in a way to process a
conversational natrural language interface other than just mapping of
phrases.  Check http://www.GyrusLogic.com who has the natural language
interface you should integrate.  Otherwise send an e-mail to
contact@gyruslogic..com
Thanks,
Peter Trompetter

> Sensory Inc, the world leader in embedded speech technologies, is now
> demonstrating its new patent-applied-for English language training
[quoted text clipped - 69 lines]
> others.  Sensory is a profitable privately held company founded in
> 1994. More information can be found at www.sensoryinc.com.
James Salsman - 18 Oct 2004 06:50 GMT
Peter,

Why has Gyrus decided to work with Java, Nuance, and SAPI, which are not
known for their small memory footprint?  Sensory can handle C&C grammars
just fine, I assure you, with much less memory overhead.

If you are interested in the consumer electronics market, I recommend
that you base your work on a low memory-overhead language such
as Tcl/Tk.  That's what I use as the dynamic and scripting language in
ReadSay's PROnounce System, which has a total footprint under 8 MB.

One note of caution:  I do not know whether Sensory's latest hardware
provides phoneme segmentation and phoneme confidence scores, but their
Fluent Speech Recognition software, which I use, does.

Sincerely,
James Salsman
www.readsay.com

> Hi,
> Great solution!  You should only make it different in a way to process a
[quoted text clipped - 78 lines]
>> others.  Sensory is a profitable privately held company founded in
>> 1994. More information can be found at www.sensoryinc.com.

Signature

www.readsay.com - maker of the ReadSay PROnounce English literacy system
 400 MHz PDA included:  $499 --  http://www.readsay.com/PROnounce.html

John Openshaw - 18 Oct 2004 08:32 GMT
>Peter,
>If you are interested in the consumer electronics market, I recommend
>that you base your work on a low memory-overhead language such
>as Tcl/Tk.  That's what I use as the dynamic and scripting language in
>ReadSay's PROnounce System, which has a total footprint under 8 MB.

Depending on your definition of embedded technologies an 8Mb footprint
isn't small either.

Regards

John Openshaw

>> Hi,
>> Great solution!  You should only make it different in a way to process a
[quoted text clipped - 19 lines]
>>> phrases so the speaker can get immediate feedback on the
>>> pronunciation. It's much like an English instructor on a chip!

Signature

John Openshaw

Mark Barratt - 18 Oct 2004 13:22 GMT
> > Peter,
> > If you are interested in the consumer electronics market, I
[quoted text clipped - 5 lines]
> Depending on your definition of embedded technologies an 8Mb
> footprint isn't small either.

It didn't use to be - but Moore's law is inexorable. I wouldn't
be surprised if it costs more today to buy a 1MB memory chip than
a 4MB one. And it seems like only yesterday that we used to dream
of affording 16kB...

Signature

Mark "Out of memory error" Barratt
Budapest

John Openshaw - 18 Oct 2004 15:30 GMT
>> > Peter,
>> > If you are interested in the consumer electronics market, I
[quoted text clipped - 10 lines]
>a 4MB one. And it seems like only yesterday that we used to dream
>of affording 16kB...

Hmm, despite Moores law 8Mb is still a bit rich for DSP based consumer
goods priced at $19...

Well it was the last time I looked...

Signature

John Openshaw

 
Sign In
Join
My Latest Posts
My Monitored Threads
My Blog
My Photo Gallery
My Profile
My Homepage

Start New Thread
Enable EMail Alerts
Rate this Thread



©2012 Advenet LLC   Privacy Policy - Terms of Use
This website includes both content owned or controlled by Advenet as well as content owned or controlled by third parties.