Dragon as speech engine-- Kickstarter

[terrydew]

Gary mentioned that to use Dragon for Voiceattack would be expensive and would require a Kickstarter find. I for one would gladly fund such a project.
Please consider and yes I realize the price of VA would go up with that feature.

Terry

[Slan]

Dunno, from a very selfish point of view, if Gary starts working on using Dragon as a possible speech engine, it means no more dev work on cool stuff

[Pfeil]

The cheapest version of Dragon Naturally Speaking 13 costs $99.99, over 10 times what VoiceAttack currently goes for.

The Microsoft Speech Recognition engine has limitations, but I'd argue it works pretty well for phrase recognition.
As it's a product that's included in the price of your operating system, you're going to have a hard time upselling to those that don't need(and I mean $100 worth of need) dictation, I feel.

[Gary]

Did I mention Kickstarter?  I just can't recall that o_O... then again I can't remember what I did this morning.  The Dragon SDK is like $5k to get on board and I'm not sure what the distributed libraries will cost per user.  There would be added cost, I'm sure.  I actually have a pending message in with the Dragon people (Nuance)... I am small potatoes and they've got bigger ones I'm sure (probably for the best).  It would be nice to find out the details at least.

Pfeil is right, though.  Phrase recognition with a trained speech engine doesn't leave much to gain by going to Dragon.  Otherwise, an alternate speech engine would already be in place.  Dictation is another story, though.  My untrained iPhone (Dragon, right?) seems to work just as well as my trained Windows engine.  They each seem to make their own mistakes, but the iPhone seems to handle contextual stuff a little better (to test I just hold my iPhone up in front my face at the same time I'm talking into my headset).

I am looking at alternatives, but I keep in mind that dictation is not the primary feature that users of VoiceAttack are looking for.  Like Slan said, I have to weigh supporting the wants and needs of the majority of VA users versus creating a feature there may be little interest in.  It wouldn't be a complete hassle to drop in a new engine, however, I'm sure there is (just as there is with the Windows engine) eight billion anomalies to sort out... and that could take a lot of time (and the pulling-out of a lot of hair).

[terrydew]

I understand the concerns over price but with VR coming in a big way there is going to be a real need for dictation as well as command phrases. I would suggest a "premium" version developed via Kickstarter and leave the current version as is. I bet that the premium version would soon take over the standard version even at a significantly higher price. I think VR is a real game changer particularly for simulations.

Just my thoughts.

Terry

[Nagual]

What are the advantages of using Dragon vs Windows? While Windows Voice Rec does make mistakes, it's not really an issue since all commands given to VA are pre-decided. Any unrecognised commands can very easily be either trained in Windows or alternative  'sounds' of the command can be inserted into VA.

Cheers.