Dragon speech recognition software is unlike any other product in the way that it requires your voice in order to navigate, issue commands and transcribe dictation. As we tell most clients, using Dragon doesn’t mean that your keyboard and mouse are suddenly obsolete. The fastest way to do something is the best way, which is why using speech recognition in combination with your keyboard and mouse generally works the best. Sometimes it’s easier to press a key on the keyboard than use your voice or to move the cursor to a different space on the document or to simply click the OK button rather than saying “click OK”.
There are cases where people who don’t have full use of their hands are actually quite grateful to have Dragon in order to make all these things possible. Generally in these situations more work is required in the way of training and customisation in order to get to a point where hands-free computing is a real possibility.
Beyond this ratio of hardware input varying between microphone, keyboard and mouse, Dragon is a living and breathing thing. We’re not talking about a fire-breathing dragon inside the software or a ghost in the machine, but we are talking about the idea that your profile is constantly evolving. Dragon’s algorithms are designed to constantly be improving in accuracy and learning from you.
Determining which words are used more frequently than others, taking note of changes you make when you are actually correcting misrecognitions and learning from your work patterns, the system is designed to adapt to you. It does require some actual input from your side when it comes to making corrections and customising your vocabulary but it’s important to keep things in check.
Your voice profile becomes so valuable over time with all the changes and customisations that it’s important to look after it. To do this you can run the Acoustic Optimizer every now and then, which streamlines the new sound information into your voice profile.
It’s also a good idea to make a backup of your user profile from time to time. If anything goes wrong with your profile this will be very useful because you’re able to actually refer back to a previous good version of your profile. There’s a chance that your voice profile can become corrupt and being able to rely on a backup just makes keeping your vocabulary intact and customised settings possible.
You can import a vocabulary to a new user profile but you run the risk of losing all the learning that it is done over the days, weeks and months been using that profile. The software is so accurate these days that it’s not as big a deal to start over but if you want to save time, make sure you have a good copy on standby.
It’s also a good idea to try and stick to using one audio input for your profile and clear out any typos or words that may have crept into the Vocabulary Editor by mistake. Switching your microphone off when you aren’t actively using the product is also a great idea to limit the recording data that Dragon absorbs.
Another tip is to watch your correction, only correcting misecognised text when Dragon makes the mistake. Being able to differentiate between this and when you make a mistake is important because it prevents you from degrading your profile over time. If you mumble or slur a word and then correct it so that Dragon thinks that’s how you pronounce that word, this will ultimately lead to fuzzy accuracy.
Doing all of these things will just assist you in getting the best possible results and having a fail-safe voice profile to refer to if you need one.