Voice recognition, text readers and sound recording.
- Speech to text software (also known as Voice Recognition or VR)
- Text read aloud
- Sound recording machines
- Personal Digital Assistants (PDAs)
- Other assistive technology
Speech to text software (also known as Voice Recognition or VR)
High tech/ high cost
The user speaks, the software recognises what was said and types it into the computer.
This means that:
- Words are correctly spelled.
- The user’s flow is not interrupted by having to stop and worry about spelling.
- The need to type or handwrite is removed: this is helpful if the user is also dyspraxic.
Dyslexia affects different people in different ways. Some people with dyslexia will be able to use voice recognition software without any problems. Others may have difficulty with enrolment, dictation or correction.
A common view on voice recognition (VR) software is that it promises more than it delivers. It can be time consuming to train and can make errors in detecting speech. That said, for students who really struggle with writing, it can be a useful tool. Improvements in the software have produced significant gains in the speed and accuracy of voice recognition, to the point where it is no longer just a tool for disabled people, but one which can deliver real benefits for anyone who does a lot of writing in their work.
What can the software now do? While you have to treat the manufacturers' claims cautiously, users can achieve very high accuracy and speed without having to spend a lot of time training the software or large amounts of cash on additional hardware.
Some useful considerations in getting hold of the software include:
Accuracy – look at how accurate the software will be (cheaper versions are not always as accurate)
Additional microphone – you may need to purchase one in addition to the software to maximise accuracy. A high quality microphone can make a dramatic difference to recognition quality.
Time – you may need to spend time teaching the software special phrases. Once taught, the software is pretty accurate at dealing with this.
Dictation style -the training materials often tell you that you should dictate as if you were a newscaster reading the news. The software won't work well if you don't speak clearly
Training the software
Before starting to use a voice recognition program you have to read out a document that is presented on the screen. This training process can be an issue for people who are not fluent readers. Ideas to get a round this include:
- Choose a suitable enrolment text: the voice recognition programs offer a choice of texts - some are easier to read than others.
- Print out the enrolment text in large print: you can familiarise yourself with the document before reading it into the computer.
- Work with a helper: the helper reads the text on the screen in small sections and the user repeats it a section at a time.
- Use an additional package like Keystone: this provides alternative training texts and text to speech support during enrolment.
- You can skip words in Dragon too.
Words are correctly spelled, but there will be some misrecognised words which will be difficult to spot:
- They will not be picked up by a spell checker: - as they are correctly spelled.
- The words will often have the same “shape”: e.g. “modern” and “modem”.
- Misrecognised words need to be corrected and this process involves choosing the right word from a list of suggestions. If the “right” word is not listed you need to type it in.
- Dyslexic users may struggle to identify the correct word in the list.
- Some will be unable to spell out the words during the correction process.
- For some users, text-to-speech software can help get round some of the difficulties of identifying and making corrections.
Using Text-to-Speech with Voice Recognition Software
NaturallySpeaking Preferred and ViaVoice have speech output facilities that will help many users. Once text has been recognised the user can:
- Listen to it using text-to-speech: you select the paragraph or sentences that have been recognised and click on the text -to-speech feature. The text is read back using a synthetic voice. Then you can follow the text on the screen as it is spoken out.
- Listen to what was said: hear a recording of your own voice – this can be helpful if they are not sure what they said.
- The text-to-speech programs provided with Naturally Speaking and Via Voice will not read out the correction lists and provide no support during enrolment
Other sources of information about speech-to-text computer software:
Text read aloud
There are a variety of different programs available on the market which read back text, but the additional features vary.
The choice of software depends very much on the extent of individual students' specific difficulties along with their preferred learning style and previous support and experiences.
Medium tech/ low cost
Read back programs are used by students to hear their own work read back. Rather than just reading over their own work and reading what they think they wrote, they actually hear the words as they are highlighted. This can often assist with identifying errors more efficiently, for example, inappropriate words, words missed out or incorrect spelling. It can also help to identify areas that do not sound right due to a lack of flow or words in the wrong order or incorrect endings. A superior spell checker, devised to help identify dyslexic spellings, also picks up more errors and provides more options than the well known Word spell checker. An additional feature is its ability to list a selection of alternative words to replace simplistic language or when word finding difficulties arise.
- ClaroRead and Word Read http://www.clarosoftware.com/index.php
- ReadPlease http://www.readplease.com/
- SpeakOut http://www.screenreader.co.uk/
- Text to speech on the web – BrowseAloud (http://www.texthelp.com/
- Read and Write (earlier versions than Gold) (http://www.texthelp.com/
It is sometimes possible to download free trials of the software from the sites listed above.
High tech/ high cost
A more sophisticated version such as TextHelp Read & Write Gold, will also scan in text and read it back. This is often very beneficial for students who find reading text a slow and laborious process, have to reread for comprehension and have difficulties with pronunciation and recognition. Often they are able to comprehend and learn more quickly if they hear the information read to them as opposed to having to read and decode words for themselves. There are many other features with this, but the reading function tends to be one of the most useful for students with specific learning difficulties.
Some examples include:
TextHelp Read and Write Gold http://www.texthelp.com/
Sound recording machines
Sound recording machines can be useful for the following:
- Recording lectures, meetings or seminars for future playback and capture
- Recording your own voice for example thoughts on an essay, before writing out, revision topics, memos or tasks.
- Additionally, you could carry one round with you to capture thoughts and ideas on the move.
Low cost/ low tech
Simple dictation machines can be obtained for around £20 (UK price, Dec 2006). These have small tapes and have the capacity to be used for simple recording of thoughts and playback. They do not have sufficient microphones to pick up sound in meetings or lectures however. To obtain one, try a high street electrical retailer or catalogue store. An example recommended by techdis.ac.uk is the Sony TCM450DV
Medium tech/ medium cost
Digital recording devices allow high quality sound capture, but are more limited in terms of file conversion or additional features. An example recommended by techdis.ac.uk is the Sharp MD-MT290H (S) Personal Minidisk £70
High tech/ high cost
Digital recorders enable higher quality sound capture, particularly if used with an additional external microphone. The cost of the recording devices is roughly £150 and around £50-£80 for a high quality external microphone.
Some recording devices are compatible with speech to text software and record in Digital Speech Standard(DSS) which can be used by Via Voice speech recognition or converted to WAV files for Dragon Dictate Naturally Speaking.
Some examples are listed below with annotations based on information available at www.techdis.ac.uk
The Olympus DM-1 digital voice recorder and music player supports both MP3 and WMA files. The recorder has an internal speaker, microphone and headphone jack and takes batteries. It has an LCD display which shows file information and recording settings and the buttons control single actions so the recorder is not entirely menu driven. There is an external power source through a connecting AC Adapter (sold separately). It is easy to hold the recorder in the palm of the hand and it feels relatively robust. It is suitable for both lectures and tutorials as well as personal notes for speech recognition. The MP3 player can store approximately 60 minutes of music on 64 MB at 128 kbps. The system provides 3D sound quality due to the WOW technology, and the player also has equalising, repeat and random play functions. This model is currently the most likely to be recommended as part of the Disabled Students Allowance in the UK
Other available models include:
- The Olympus DS-330
- The Olympus DS-3000
- Sony mzb10
Please note, that not all are compatible with Dragon and Via Voice software. It is advisable to think carefully about how you might use a dictation machine as there are several hundreds to choose from and to carefully look at the features, prices and designs of the available models.
Personal Digital Assistants (PDAs)
A PDA is an electronic personal organiser. They can be a good way of overcoming some of the organisational difficulties associated with Dyslexia. There are lots of different models available. There are two main operating systems that dominate the market (Palm OS and Windows mobile) For more information on the pros and cons of the Operating Systems, see www.dyslexic.com.
Medium tech/ medium cost
Just about every PDA on the market will have the following characteristics and features:
- Set reminders and tasks
- Note taking
High tech/ high cost
Some more expensive models have additional features, for example:
- Voice memo
- Transcriber/Graffiti (handwriting recognition)
- Keyboards (folding standard QWERTY, virtual ones that use laser technology or thumb (like mobile phone pad)
- Synchronisation - keeps information current on PC and PDA; when you synchronise your PDA with your computer the two machines communicate and swap information so that both have the same current information. Synchronisation is also the way to install new programmes and move files backwards and forwards between the PDA and the Desktop. Most PDAs have the ability to synchronise:
- Word docs
- Reminders and tasks
Software available for PDAs includes:
- Text to speech
Other useful features available in the more expensive PDAs include:
- Wireless internet access
- Satellite navigation
- Music and video playback
As an alternative to a PDA, you may also wish to consider a Smartphone. These are new generation mobile phones with additional features.
Other assistive technology
There is a huge range of support software and devices available to support individual needs, for example trackerballs and ergonomic mice and keyboards for the computer and typing tutors.
There are two good sources as follows that aim to provide information about the resources available:
www.dyslexic.com (this includes product comparisons and reviews)
McLoughlin et al (2002) The Adult Dyslexic: interventions and outcomesLondon: Whurr