April 2002 NGTVoice

  1. DNS 6 Acoustic Optimizer
  2. XP Restore Utility
  3. CSUN Conference Update
  4. Telex Broadcast Systems for the Classroom
  5. Speech Technology Forum Created
  6. GN Netcom's 8050 Switch

DNS 6 Acoustic Optimizer

I recently had the opportunity to explore the advantages of running the Acoustic Optimizer on a biweekly basis. I'm suitably impressed.

Info from DNS Help File

As you make corrections and perform additional training, Dragon NaturallySpeaking stores this acoustic data in an archive and uses it to enhance recognition accuracy. The Acoustic Optimizer accumulates this data from session to session. This has several advantages over normal adaptation and correction:

To run the Acoustic Optimizer

  1. On the Tools menu of the DragonBar, click Accuracy Center.
  2. In the Accuracy Center click or say, Run the Acoustic Optimizer.
  3. On the Acoustic Optimizer dialog box, click Go.

The Acoustic Optimizer may ask if you want to save your current user files. If the corrections you have made during the current session have improved your accuracy, click Yes.

The Acoustic Optimizer dialog box will display the amount of time to finish adapting your user files and display a progress bar to indicate how much of the job is complete.

XP Restore Function

One of the nicer features of the new XP Operating System is the ability to capture and store computer system images for future restore activity. This utility should be a godsend, but it needs to be initiated in order to capture a system image. We've started shipping our NGT Voice Pro systems with an initial system image captured on the users hard drive. Here's how to use this utility: use the Start Menu/Programs/Accessories/System Tools/Restore. Follow the Wizard to create a restore point. In future, if you need to restore your system to the previous image follow the same path and select restore.

CSUN Conference Update

Ed Rosenthal recently returned from the CSUN Conference on Technologies and Persons with Disabilities. In addition to a six-hour technical training session conducted on March 18, an informational one-hour session was also conducted which was attended by approximately 115 people. The conference was significant for number of developments; one of which was the attendance of many federal/state IT professionals seeking information about accommodations for people with disabilities focusing on technology. Section 508 compliance issues may drive some of this, but one other underlying issue seemed to be the general improvement in adaptive technologies in general.

Telex Broadcast Systems for the Classroom

For several years we've been working with the Telex microphone/switching equipment line. The people at Telex have just introduced a new line of cordless microphones and broadcast speakers for use within classroom environments. We've had several occasions over the last few years where we've consulted on broadcast systems for classrooms. Now that we have one that is so configurable and well-designed, I think we'll see more call for this type of accommodation. One of the next areas to explore would be to see if we could use the same wireless broadcast microphone worn by the instructor to send a signal to a speech recognition desktop/laptop computer at a students workspace. This would allow a student who is hard-of-hearing or deaf to participate in classroom without having to hear what the teacher is saying.

CharacTell's iREAD Notes Pro

Most of the details associated with the new iREAD Notes Pro handwriting conversion technology have been worked out. And I anticipate being able to ship products towards the middle of this month. In the March 2002 Advanced Discussions newsletter a full description of the capabilities of the proprietary Advanced Character Recognition technology (ACR) is provided. Here's some basic information Currently, handwriting recognition achieves recognition rates of approximately 85% for mixed operand lowercase alphabets (non connected) and about 90% for upper cases only. This means that almost every second word requires human correction. Meanwhile, research has shown the people find handwriting recognition useful only when it reaches rates of over 97% of characters. ACR, in most cases, offers success rates for upper and lower (non cursive) characters around 98%, and for uppercase only, as high as 99%! We're excited to be bringing this revolutionary technology to North America.

Speech Technology Forum Created

A new forum has been founded to improve business understanding, acceptance, and development of speech technology platforms, tools, and applications. according to the latest issue of Speech Technology Magazine. This forum is based out of Texas, and has a number of high-profile participants. The trend of this certainly coincides with what we've been seeing over the last few years with increasing acceptance of speech recognition technology as a primary input tool and, the desire for organizations to integrate speech recognition directly into their proprietary applications and workflow.

GN Netcom 8050

GN Netcom has recently announced the replacement of its MPAII/CT Switch combination that has been used very successfully by our clients to use a single headset to speak to both telephone and computer system. The new replacement is called an 8050 universal amplifier/computer switch and we've begun testing in our labs. Intended to be used with the OG 1 /OG 2 corded headset microphone we expect smooth integration with speech recognition applications.

Best wishes for a safe and happy spring season!

Ed Rosenthal
And the staff at NGT

Page Last Updated: Sunday, June 01, 2008

Copyright 1999- 2008 Next Generation Technologies Incorporated

Jump to: site navigation | page content

Highlighted Products and Services

Site Navigation