Voice Recognition Software Reviews

A recent website visitor asked the question…”Is voice recognition software worth the investment?” The answer is, it all depends on your usage of such software, the environment you will be using it in, and how much patience you have to properly set it up.

Voice recognition, or automatic speech recognition, is a technology dating back many years. Webopedia defines voice recognition as The field of computer science that deals with designing computer systems that can recognize spoken words. Note that voice recognition implies only that the computer can take dictation, not that it understands what is being said.

When voice recognition first became available, it primarily benefited those with writing and learning difficulties. One possible reason why such technology did not hit mainstream consumer markets was that it initially was ‘discrete-based’ recognition only, which means you had to pause the computer before moving on, as it could only provide possible answers.

Thanks to advancements in speech to text technology, modern voice recognition applications use what is called ‘continuous speech technology’. This allows the user to speak more naturally, while the computer dictates what it hears. Pending the speed of you computer, you may find such software having to ‘catch up’ to your voice. The accuracy of the software is truly based on the amount of ‘training’ you provide. Training the computer to recognize your speech patterns and inflexion’s will help increase accuracy by building a larger vocabulary to reference.

This article will cover the benefits of, how to choose, and factors to consider as you compare various voice recognition software brands.

It may be worth mentioning at the start, that I am a bit biased towards Dragon NaturallySpeaking, by Nuance. I have tried several brands, and although it is more expensive than most, you do get what you pay for.

Benefits of Using Voice Recognition Software

Okay, let’s be straight…voice recognition software is not AI (artificial intelligence); it’s not on the level as speaking with the Star Trek ‘Computer’. However, the more training you do with the application, the more accurate it will be in understanding your voice.

Assuming a quiet environment, or the use of a noise-resistant, or noise proof microphone, the biggest benefit of such software would be to allow you to operate your computer almost hands-free. Along with touch-sensitive LCD computer screens, you would barely have to operate a mouse or keyboard.

Though the learning curve may seem a bit intensive, using your voice for commands is by far faster than any combination of keyboard and mouse instructions. Once you get familiar with the various voice commands of your application, you will find it easy to navigate your computer, dictate notes, write articles, and surf the net. As a matter of fact, I’m using Dragon NaturallySpeaking to write this article.

How to Choose The Best Voice Recognition Software

In this software market segment, it is more about the performance and operation of the software. Basing your purchase on price will not yield you the best results. With most software applications, if you purchase a less-expensive brand, you’ll probably end up with less features, though the main functionality will be the same. If you were to purchase voice recognition software in the same manner, accuracy and functionality will be limited.

Your best bet is to strive for a product that has no less than a 90 percent accuracy mark. I can attest that with proper training, DNS is about 99% accurate for my use. Additionally, you want a program that has a good corrections ability, and flexible word database. Once that allows you to quickly train and add words to it as you move along in your work. This may seem intensive initially, but the more you use the software, the less training it will require.

What to Look for in Voice Recognition Software

Here’s a good list of what to look for in making your speech to text software selection.

Compatibility

Make certain your computer can handle the voice recognition software you wish to purchase and install. For me, Dragon NaturallySpeaking requires approximately 300 MB of RAM to run, and recognize it takes a good chunk of CPU cycles to operate smoothly. I have found other brands to not take as much in resources, but expect to have limited functions, or options. Here’s the requirements for DNS: Dragon NaturallySpeaking

Check if the software was developed to work with your software. DNS has been able to work well with most of the software I have. A good example of this would be this issue covered by Nuance Support: Unable to access JAVA applet pop up menus by voice. The solution provided in Tech Note ID 4140: You can navigate to these buttons using the “MouseGrid” or “MouseGrid Window” command. When the menu pops up, you will have to leave the MouseGrid context by pressing the escape key. The menu then stays available and you can say any of the items in the menu to go to where the menu item points. There are a couple more options mentioned as well.

Always a good option would be to try Windows built-in voice recognition commands. Read this article for assistance in using: How to Use Speech Recognition with Windows 7, at Windows 7 News & Tips. The author says, Windows 7 has many features designed to let the user interact with the desktop interface in place of the keyboard and mouse. GGarza.

Usage

Is such software conducive to the type, and style of writing you do? An example would be writing computer programming code, or software engineering. There are so many variables, acronyms, and non-traditional words used, it takes a fair amount of time to train the recognition software.

A similar question was asked on stackoverflow.com: Speech Recognition & Programming. The question asked was, Has anyone had success with Dragon Naturally Speaking voice recognition software when it comes to programming? There were several answers including, I think that “voice programming” and “programming by voice” search better “speech recognition programming”. It has been tried but not yet caught on. By dmckee.

Pay some attention to various editions within each brand as well. For example, DNS Home Edition does not work with Microsoft Excel right out of the box.

Last, think about the environment you are in; is it quiet, very loud, etc? If it’s quiet, will you be disturbing others? If you’re in an office, you are probably good to go.

Accuracy

As previously mentioned…check the accuracy of the software. Time permitting, browse forums and answer keys such as Yahoo Answers for feedback on various brands and versions.

Is one thing that many skip when they start comparing prices…the method of input. Quality of microphones make a big difference. A good microphone allows clear input, which allows the software to match the sound with a better match the database more quickly. A good microphone came with the purchase of my DNS Premium. On a side note…I tried using my Bluetooth headset with DNS, and although it worked, it keeps the headset active, thereby, draining the battery quickly; not recommended.

Accuracy can really be graded in using the dictation function of the application. How well does it recognize your voice, after appropriate training?

Commands and Functionality

The primary function of the software depends on your usage and the brand. Regardless of the use, the application must have a good voice recognition database to match commands with. Also, it should have an ease to follow manner of training and adding new words to the database on the fly.

Features

Often, we get fixated on how many features a particular software program has. For instance, in my personal opinion, Microsoft Word has way too many features and options. The use of can really decrease the productivity of an individual. With that said, just make sure the program comes with the proper features needed, such as voice training modules, and database access features.

Summary

In summary, you can see that it does take some effort to make any voice recognition software to work smoothly. From personal use, I can say that it is very effective, once I trained the software to recognize more words and commands.