Dragon voice recognition resources and news

Voice recognition tips to enhance your computer productivity……….

A powerful tool to increase accuracy

Posted by Peter Maddern on May 26th, 2009

It can be extremely worthwhile running the Acoustic and Language Model tool in Dragon NaturallySpeaking on a regular basis. This is a valuable tool to increase your accuracy from your cumulative corrections (as long as you have diligently been using Dragon to make corrections that is!)

What is this tool and what does it do?

  • The Acoustic and Language Model Optimizer can automatically “tune” Dragon by making it review all the acoustic data and/or text data it has gathered from your usage. You may wish to run it after a few weeks of using Dragon regularly, and subsequently about once a month or even more often.
  • This Optimization can be launched at any time from the Accuracy Centre. It can also be scheduled so that Dragon performs it at a specific frequency and time.

You can do this manually or for convenience, Dragon lets you schedule a specific time and frequency for these optimizations.

To launch the Acoustic and/or Language Model Optimization manually:-

Step 1: In the Accuracy Centre, use the link “Run the Acoustic and Language Model Optimizer”
Step 2: Check the Acoustic or Language Model checkbox as desired, then click Go. (As the dialog box mentions, you should run this tool when your PC is not being used actively.)

image To set up an automatic schedule:-

Step 1: In the Accuracy Centre, use the link “Schedule Acoustic and Language Model Optimizer to run later”. This opens the Accoustic and Language Model Optimizer Scheduler.
Step 2: Select the user profile from the list of profiles presented
Step 3: Go to the menu bar and click on “File” followed by “New Task” – you will be prompted for your Windows user name and password
Step 4: Follow the prompts to enter the time for optimization to take place. Choose a convenient time when you don’t actively use the PC e.g. I choose lunch time

See also this previous blog post of mine on the subject.

Peter Maddern

Posted in Dragon NaturallySpeaking tips | No Comments »

New Resources on Using Dragon NaturallySpeaking

Posted by Peter Maddern on May 26th, 2009

492740_library_books Nuance have updated their website with some extremely useful resources which can help you better use Dragon NaturallySpeaking.

This page hosts the Dragon NaturallySpeaking 10 end-user workbook, a section on customising the Dragon NaturallySpeaking vocabulary and a “cheat sheet” of key tips and commands.

This page hosts the Dragon NaturallySpeaking 10 user guide. Each section is available in document form as well as a playable recording (wav file).

It’s well worth printing out and reading these resources in detail.

Peter Maddern
Speech Empowered Computing

Posted in Dragon NaturallySpeaking tips | No Comments »

Follow Speech Empowered Computing on Twitter

Posted by Peter Maddern on April 24th, 2009

image By now, you must have heard of Twitter? Think of it as a micro – blog. You send maximum 140 character updates about what you are doing and you follow updates of others you choose to follow and others may choose to follow you. It’s quite addictive!

You can follow Speech Empowered Computing at

www.twitter.com/speechem

Peter Maddern

Posted in Speech Empowered Computing | No Comments »

Dragon NaturallySpeaking 10.1 for Vista 64 bit

Posted by Peter Maddern on April 23rd, 2009

Since the launch of version 10 in September 2008, Nuance has been collecting your feedback and working on improvements to make this new version perfect to suit your needs. They would normally wait for version 11 to launch a new version, but the improvements they’ve made to this version are so important that they apparently wanted to share them as soon as possible.

Works with VISTA 64-bit

Read the rest of this entry »

Posted in Dragon NaturallySpeaking software | 3 Comments »

How does speech recognition work?

Posted by Peter Maddern on April 19th, 2009

Speech recognition as exemplified by Dragon NaturallySpeaking is a complex process but can be broken down to contributions from 3 models as follows:-

image 
The “Acoustic model” represents the connection between language sounds (phonemes) and the voice signal.

The “Lexicon” represents the connection between phonemes and words.

The “Language model” represents the connection between words and phrases/sentences.

Peter Maddern

Posted in Speech recognition | No Comments »

Check your audio settings frequently to maintain your accuracy

Posted by Peter Maddern on April 19th, 2009

Tip - Check your audio settings in Dragon NaturallySpeaking on a regular basis.

Now and again, you might find that your accuracy diminishes. Read the rest of this entry »

Posted in Dragon NaturallySpeaking tips | No Comments »

Want to help make Dragon an even better product?

Posted by Peter Maddern on April 10th, 2009

The product marketing manager for Dragon NaturallySpeaking is asking for your help regarding future versions of Dragon NaturallySpeaking. For the past few months, Nuance’s product marketing manager has been reaching out to present and former Dragon NaturallySpeaking customers for product feedback. Users’ opinions will enable Nuance to fine-tune the specific areas of the product that you would like to see improved. For example: Read the rest of this entry »

Posted in Dragon NaturallySpeaking software | 1 Comment »

New screencast demonstrations

Posted by Peter Maddern on April 7th, 2009

I have made some new online demonstrations of two Bluetooth headsets.

See:-

Sennheiser VMX 100 Bluetooth headset

Plantronics Calisto Bluetooth headset (the headset in Dragon Preferred and Professional Wireless)

Peter Maddern

Posted in Microphones | No Comments »

Selecting groups of cells in Microsoft Excel

Posted by Peter Maddern on March 31st, 2009

When using Dragon NaturallySpeaking in both Microsoft Excel 2003 and 2007:-

To select 2 adjacent cells, select the first cell and then say “press shift Right Arrow “

To select 2 adjacent cells in the next row, you would then say “press shift down arrow”

To select a group of cells – say I wanted to select cells A2 to E6 in the following example:-

clip_image002

The command would be “select cell A2 to cell E6” or “select cell A2 through cell E6” or better still, use the international communications alphabet and say “select cell alpha two to cell echo 6”

If you want to practise using this command, I’d suggest writing the command down beforehand to read from as it can be a bit of a tongue twister!

Peter Maddern

Posted in Dragon NaturallySpeaking tips | 2 Comments »

MacSpeech Dictate for Intel Macs now available

Posted by Peter Maddern on March 23rd, 2009

MacSpeech

Speech Empowered Computing is now pleased to add MacSpeech Dictate for Intel Macs to its range of speech recognition products.

MacSpeech Dictate Regular and Education are available from this link.

Training is also available in MacSpeech Dictate.

Contact Speech Empowered Computing for more information.

Peter Maddern

Posted in MacSpeech | No Comments »