Back

Voice input

Introduction
Using voice as an input method has become increasingly popular, although there are still problems associated with using this methods for input.

micTypical uses
Voice recognition software is a program that accepts your voice via a microphone to allow you to enter commands or to allow you to enter data. There are lots of different uses of voice input. 

Information systems
Information systems can be used that accept simple commands from a user and then gives them back information. For example, if you ring up a bank, they may have a system that asks you what you want. They will give you some ideas of what to say, for example, 'PIN', 'Balance', 'Mortgage', 'Savings and so on. Each answer is analysed and will usually lead to a further menu to narrow down your requirements. Eventually, you will get to a recorded message giving you the information you need, or you will be put through to a person who can help you. You can often phone an airport and will be greeted with an automated information system, where you might be prompted to say, 'Arrivals', 'Departures', 'Parking' and so on.

Steering calls
Many companies use a system where you are asked what your query is about, then you are taken through some security details and then you are transferred to the right operator. This system is often found in any company where you have an account, for example a utilities company or a bank. 

Biometrics
Companies are increasingly using a technology that makes and keeps a 'voice print' of your voice as a security measure in the never-ending fight against crime. The next time you phone in, you simply speak your name and it is compared to the voice print the company has of you. Once you have been authorised, you are then put though to a person who can deal with your query.

Data input
It is often quicker, especially if you don't have typing training, to speak a letter or email rather than type it in. There is usually a learning process that you have to go through with the software, to reduce the chance of incorrect entries, but once this is done and you have learnt how to manipulate the voice recognition software, it can be very accurate. Even with strong accents or voice changes due to a cold, voice recognition software can be very powerful and improve productivity.

Voice commands
Many computer systems including phones and tablets have built-in software that allows you to use your voice to instruct the device to carry out basic tasks, such as opening and closing an application, doing an Internet search, saving a file you are working on, deleting a paragraph and adjusting the volume, for example. Again, there is normally some training involved, where you have to speak commands into the software and the software' learns' your voice, but once this has been done and you have got used to the commands to use, it can speed up using the computer. This can be particularly useful for people with disability issues.

Advantages and disadvantages of voice recognition

    • Speech is a very natural way of interacting with your computer or other computing device. Sometimes, you don't even need to be sitting at the computer in front of a keyboard - you could be walking around talking into a Bluetooth microphone linked to your computer, for example.
    • You have to teach your software to recognise your voice, to improve the accuracy of the software. To do this, you have to speak in words so that the software can start saving information about your accent, how you say words, your pronuciation, the speed that you speek at, the volume level of your voice and so on. This can take time. 
    • The software knows that mistakes are going to happen from time to time. When a mistake happens, it will often display a number of close matches for the user to select from. This can slow data input down, depending on how accurate the software is and how much time you have spent teaching the software about your voice.
    • The accuracy of the software can be reduced if your voice changes due to a cold or there is a lot of background noise, for example, from a TV, a radio, the phone ringing, a kettle boiling, a pet making noises, the doorbell ringing and so on.
    • Programs cannot understand the context of a command in the same way that humans work out the context from what you said and how you said it. This can lead to problems if you use slang, or for homonyms (words that sound the same but are spealt differently, such as 'their, they're and there' or 'two, to and too').
    • Although voice recognition systems can speed up work, you have to invest the time teaching the system and correcting errors. Many people find that this is a frustrating experience.
    • You have to buy a good quality microphone for the best results and commercial voice recognition software can be expensive.

Back