Keystone SpeechTutor
My Shopping Cart
You have 0 items in your cart.
TOTAL: £0.00
(Excl. VAT & Del.)
Register | Login
Products
Special Offers
Meeting your Needs
Reviews
Presentations
User Guides
Product Categories
Literacy aids
Voice Recognition
Services
Literacy support bundles
Microphones and headsets
Complete workstations
Mind Mapping
Dictation Units
With no ties to any supplier, we can provide objective support & guidance.
Search Words Worldwide:   or

Literacy aids - KeyStone SpeechTutor 3

Store > Literacy aids > KeyStone SpeechTutor 3

KeyStone SpeechTutor 3


KeyStone SpeechTutor 3
£105.00 (Excl. VAT)

Download the free evaluation system to try out KeyStone SpeechTutor.

KeyStone SpeechTutor is a unique product. Its main function is to provide a platform to enable novice users of speech recognition to become familiar with the techniques. It also assists in understanding how misrecognition can occur and how to correct for this. Furthermore, it provides an objective means of testing microphones and the recognition process itself.

Download the PowerPoint presentation  to get full information on this exciting product.

The program can be used with Dragon NaturallySpeaking (versions 8, 9, 10 and 11), DragonDictate (version 3) and the Microsoft Vista voice recognition system.

It has four modes of operation.

Tutor Mode allows users to practice using the technology and assesses their performance in an objective manner. Scores are calculated by comparing the results of users speaking short sentences which are prompted by the computer. Statistics are collected during each session and results sheets can be retrieved by the teacher or the user

This simple incentive scheme encourages users to improve their enunciation which leads to higher recognition performance and a greater success rate in using the technology. Their efforts are assessed automatically so that the teacher gains information on where difficulties occur.

Typical users of KeyStone SpeechTutor are those with speech production problems, those who are young or inexperienced and those whose first language is not the language being used for dictation.

More detailed information on the unique automatic coaching system is shown below.

Active Mode is particularly useful for those who find it daunting to use speech recognition with a word processor.  The special text window is simpler to use than Dragon NaturallySpeaking's DragonPad. Restricting the activity to a simple environment rather than using a word processor environment like Microsoft Word is also of great assistance.

The playing back of dictated text and reading out of text by the text-to-speech system simplifies these activities. The user is able to hear what has been spoken and also what has been produced on the screen by simple button clicks.

Finally, the dictated text can be scanned for words which are often incorrectly recognised, particularly homophones such as "road", "rode" and "rowed". These can then be reviewed and corrections can be made where appropriate.

Analysis Mode enables trouble-shooting to be carried out. An objective scheme allows objective tests to be made of different microphones, operating conditions or users for fixed sets of word groups. Statistics and results sheets can be retained for checking purposes.

Enrolment Mode provides a special environment whereby a novice user can enrol more easily than through the normal scheme. Each short sentence or phrase is prompted for the user by a computer-generated voice. This also assists people suffering from visual difficulties.

Student progress monitoring can be carried out so that the user's performance can be checked and the sources of any trouble can be identified.

Download the PowerPoint presentation  to get full information on this exciting product.

Download the free evaluation system to try out the product.

SpeechTutor is also provided within the KeyStone SpeechMaster bundle.

SpeechTutor is compatible with Microsoft Windows XP, Windows Vista and Windows 7 operating systems. 

 

The automatic coaching system: background

Speech recognition technology has advanced greatly in recent years. Many people with quite severe literacy difficulties now use the techniques to assist them in a variety of ways. However, there remain many who simply cannot make sufficient progress to justify continuing in the attempt.

Some users find difficulty in adjusting to the idea of speaking out their thoughts. Others experience problems with enunciation and pronunciation.

KeyStone SpeechTutor was introduced to address these problems and has proved to be a powerful tool. It enables users who have successfully enrolled using Dragon NaturallySpeaking or DragonDictate voice recognition products to practice speaking, being prompted by a computer voice.

A professional assessor can use KeyStone SpeechTutor to check how well the speech system responds to a candidate's voice. The assessment environment is isolated from any other activity on the computer, so there are no complications which could lead to poor recognition causing unexpected and confusing events.

However KeyStone SpeechTutor has also proved to be ideal for taking a novice through the early stages of using speech recognition. The teacher can concentrate on improving the trainee's speaking style, correcting the speech recognition system as necessary.

Once the mechanism for producing text is understood and the user has started to build confidence in this way, the process of learning to use voice recognition in earnest can begin. Only then should steps be taken to teach the user how to correct mistakes made by the speech recognition equipment.

A group of "lessons" are available within the program, each of which consists of a number of carefully prepared short sentences. Different groups of sentences can be used in separate lessons, each concentrating on specific difficulties being experienced by the trainee.

The assessor selects an appropriate lesson to match the candidate's circumstances. For instance, if difficulties are expected with words starting with "th", a lesson would be selected containing several sentences such as "This is thirsty work".

Sentences are displayed successively and the integrated text-to-speech system is used to prompt the candidate, who repeats the sentence. The result is scored by a machine comparison of the resulting text with the original prompt. A perfect match produces a 100% result and the next sentence can then be selected.

After an unsatisfactory attempt (showing a score of less than 100%) the assessor can carry out one of three possible actions:

1    If the attempt was reasonable, the speech recognition system can be instructed to accept it and to adjust the stored statistics relating to the user's voice. A new attempt can then be made to check whether the adjustment has been successful.

2    The candidate can be asked to repeat the sentence to see whether improvement in enunciation is feasible.

3    The example can be skipped and the next sentence is then displayed.

In this way, the assessor can quickly draw a conclusion as to whether the candidate has problems in enunciation or the equipment is operating less than effectively for the particular individual. In either case, a definite strategy can be drawn up to address the issues.


Stock Status: In Stock
online casino