This morning, it’s getting even better. Swype has just launched a new Beta with two fancy new tricks: Dragon-based voice dictation, and an improved ability to analyze the words you’ve already typed to better determine what you’re trying to type next.
While Swype has had voice dictation features for a while now by way of the system built into most Android handsets, this new system is powered by Nuance’s Dragon technology (as you may recall, Nuance purchased Swype for $100m back in October) Curiously, it seems the new system supplants the old one entirely — there’s no way to switch back, at least as far as I’ve seen.
The timing of the switch is particularly interesting, as the latest version of Android (Ice Cream Sandwich) introduces near-realtime transcription, with words streaming onto the display as the speaker speaks simultaneously — something Dragon definitely doesn’t do. Still, the company seems sure that Dragon’s offering outshines anything Google has to offer.
The Beta’s other new trick is called “advance language modeling” — which, as mentioned above, analyze’s what you’ve typed so far so that it’s got a better understanding of what’s coming next. To snag an example from the demo video below: take the term “Mosh pit”. You drag your finger through “M-O-S-H”, it knows you’re trying to type mosh. But when you drag through “P-I-T”, you’re also hitting “P-O-T”; do you mean Mosh Pit, or Mosh Pot? Probably the first one, as the second one isn’t a thing. If you’d instead typed “Mush” as the first word instead, it’d know you probably meant “Mush Pot”.
If you’re already in Swype’s Beta program (or are looking to get in), you’ll be able to find all of the details right over here… eventually. While the Beta has just launched, it seems their servers are currently having trouble staying up.
TechCrunch
0 comments:
Post a Comment