In December, 2003 the Boston Voice users group (BVUG) and its New York City counterpart (NYPC) did top 10 lists of what they would like to see in speech recognition engines. At the time both Dragon NaturallySpeaking and IBM’s ViaVoice were available.
Here’s my version for Dragon NaturallySpeaking 10. This list is also posted on the UC Exchange Wiki so I can keep track of whether and when they’re implemented.
1. I’d like a default user option that would let me start the program hands-free.
2. I’d like the ability to check audio settings hands-free.
3. I’d also like ability to save and switch Check Audio settings — this is useful if you travel a lot. I do an audio check whenever I land someplace new, but there’s no reason I should have to do another audio check rather than go back to a saved once I’m back in the office. I have a couple more minor suggestions for the Check Audio dialog box. First, it’s important enough to deserve its own menu item rather than only being buried in the Accuracy menu. Second, there’s an interface gotcha. Once you’ve finished checking the microphone, the focus is still on the go button. If you’re not thinking and click without moving the focus you find yourself checking the microphone again instead of going onto the accuracy check, which at best makes the process longer, and at worst is confusing.
4. I’d like separate the controls for buttons and menus. I’d like to be able to say whatever’s on the button — “yes”, “no”. But at the same time I want a longer command for menu items, e.g. “File Menu” rather than just “File”, because menu options are often active when I’m writing text.
5. The Dragon NaturallySpeaking engine should understand that when I say “Cap” what I’m looking for is a written word, not a number or symbol. “Cap Sixty” should return “Sixty”, not “60”. And “Cap Ampersand” should return “Ampersand” not “&”.
6. In the Spell Correction dialog box, I’d like a way to tell NatSpeak to type a a whole word. I’d like to say the word “Word” to indicate that the rest of the phrase is going to be a word just like I can say “Spell” to indicate that the rest of the phrase is going to be spelled.
7. The old Dragon Dictate where you could say separate words was better for people who have some types of disabilities. Putting a “speak words separately” mode in NaturallySpeaking would help a lot of people.
8. I’d like the option to be able to train the NatSpeak speech engine by repeating audio read to me through headphones rather than reading from text. This would also make training easier for younger kids.
9. I’d like a simple way to duplicate a user. Right now you can do this, but it’s a multistep and confusing process. To make a copy of the current user you have to backup, then restore. A separate menu item for duplicating would take the confusion out of the process.
10. Bring back the Dragon logo:-). The Dragon was much cooler then the green spiky blob.
What do you think of my top 10 list for NaturallySpeaking? What’s yours? Reply here or let me know at info@ this website address.