Aerofly FS2 + VoiceAttack = Awesome!

So, for months I've been using VoiceAttack with Elite Dangerous, and have thought for quite a while that especially in VR, VoiceAttack would be a dream come true for immersive control in the cockpit, without having to peek and access the keyboard or work with sometimes fiddly mouse interaction.

Several times I wanted to ask the developers about working with VoiceAttack, but I'm aware they don't really want to bring in outside programs, so I figured it would end up being very low priority. (If there was any interest at all)

Enter this guy, my new BFF!

I'm hoping this vid will stir some interest!

Updated Profile

Quote from Jet-Pack

My only concern with any speach recognition software is that you sometimes have to train it before you can use it. And that, for me, is a no go. It should work out of the box without using Microsofts inbuilt speach recognition where I would have to sit there and read text in a "foreign" language. So yeah, it should understand me even if I use German commands or mix it with English commands.
Cheers,
Jan

There's an old saying about not letting the "Perfect" be the enemy of the "Good enough".

I don't think there are any flawless speech recognition systems, but do I think the success of VoiceAttack has proven it to be good enough. One thing I have found though with these systems is that the hardware is very very very important.

I purchased a low priced headphone with mic, and couldn't really get voice recognition to work at all. It was a very very frustrating experience, and windows misunderstood me again and again, making me come to the conclusion that the software simply wasn't ready for primetime.

Later, using the apparently very high quality rift Microphone, I found that VoiceAttack suddenly understood me almost flawlessly. Belatedly I realized that not all mics are made alike! And usb mics are apparently the best choice.

Here's a video of Voiceattack working as intended in Elite Dangerous: There are also vids of it understanding other languages, but I suspect strongly things like this will have to be done by third parties. (And I bet it sells like hotcakes)

It is nice to see a proof of concept in Aerofly, though.

[video=youtube_share;s4QkltClArw]

I think that profile will become very popular! It remains to be seen whether Mike will continue to add features, or if this might become a crowdsourced effort, expended upon by the community.

If I were him, I would consider cleaning it up, expanding and documenting it, then putting it up for sale.

LOL. I've posted about it several times, but that was on Avsim.

The videomaker has posted a link. Just look under "show more" on his video notes.

Quote from Jet-Pack

Google's speach recognition works out of the box. I can ask it to get the weather for the next day and it searches for it automatically. And Apple's counterpart "Siri" also does not need any training by the user. Both of these are probably "trained" in advance to understand more people out of the box.

They also send your queries out to a Darpa/Calo AI machine-Learning type supercomputer to figure out what you just said. :cool:

Neither work without an active online connection, because the devices they operate on simply don't have the onboard computational power to parse speech that accurately on their own.

Its kind of an apples to orangutans comparison, because of that.