Speech Recognition Software and Its Importance
Whether we think we know what speech recognition software is, or not, the chances are good that we all have interacted with this software at one time or another. It is often used in telephony systems, in order for a particular system or application to be controlled via spoken commands. The voice dialing function on your mobile phone is made possible by speech recognition software.
The terms speech recognition and voice recognition are usually considered interchangeable, but technically they are not. Speech recognition is designed to recognize individual words, and perform a specific action programmed into the software when it hears that word. Voice recognition, on the other hand, is more comparable to fingerprint recognition, iris recognition etc, and is more correctly applicable to security applications. Every person has a distinctive voice and voice recognition is more about distinguishing these differences. It can be used in conjunction with speech recognition as a security measure.
As an example, a sample of your voice may be stored in a voice recognition database, and certain levels of authority are assigned to your voice print. When you say a word, or phrase, your voice is checked against this database to confirm your identity. Only if your identity can be confirmed, is the spoken word or phrase acted upon. If, for example, you have said “open” to unlock the door of a certain facility, the security level assigned to your voice in the voice recognition software is checked. Only if you are authorized to gain access to that particular facility does the speech recognition software come into play.
Once your identity has been confirmed, the word “open” is passed to the speech recognition software. It is the speech recognition software which knows to unlatch the electronic lock when that word is passed to it. If you had said, for example, “let me in”, the speech recognition software would not react, as it has not been programmed to perform any function when given that particular phrase.
Speech recognition is most commonly used, by computer users, as a means of activating programs on their computer via spoken commands, and for spoken data entry such as writing letters or replying to emails.
Let us look at some of these speech recognition systems.
Dragon Naturally Speaking – Home Edition
The new Version 11 is 15% faster than the previous version and includes a number of new voice commands making it easier to work with MS Word and your email. The new version also supports MS Office 2010. Dragon claims 99% accuracy “out of the box”, meaning there is little a need to “train” the software.
Dragon Naturally Speaking – Premium Edition
The “big brother” to the Home Edition, the Premium Edition is up to 100% accurate, meaning there is no need for training at all. It is also compatible with a range of wireless and Bluetooth microphones, certain voice recorders, and has great proofreading tools. Your typed document will be read back to you by a synthesized voice.
This system requires quite a bit of “training”, but voice profiles for multiple users may be set up on the same computer. Marketed as more of a voice activated interactive solution, it will let you browse the web, read your emails, play your music and video collection and more, all the while reading the content back to you.
Developed specifically for the Mac user, it has Microsoft Word, Excel and the rest of the Microsoft applications. This software is based on a licensed version of the Dragon Naturally Speaking voice engine bringing Dragon functionality to Mac users.
How many people even realized that, in Windows XP SP2, when combined with Office X SP1, users had access to free speech recognition software?
This has been vastly improved in Windows 7, and is also easier to activate and set up. It may not be as accurate without “training” your computer as the Dragon products, but you already have it. If you are not using speech recognition regularly, or just want to see what it can do for you, then try the Microsoft offering.