Collaborative, open research
06/07/2009 at 4:47 pm inglisjane 2 comments
Sound Barriers is a research project that we’ve had under way for a month or so. We hope to answer questions about voice as a user interface. Has this most natural of UI’s finally come of age? If not the when and why not. What does this mean for users, and user experience design. How will this impact product design and development management. Focusing on mobile apps and devices, we’re looking at applications you can control using any voice (so not voice tagging).
The project is not for a client, it’s just born out of interest and a belief that increasingly we will be working with voice as an option or as part of richer application interfaces from devices. So I am carrying out the research using as many ‘free’ means at my disposal. I also know this is a vast and complex field, so interviewing experts from the voice ‘eco-system’ as well as those in mobile, is only part of the process. I’d also like to engage in debate online here through the blog. Working in this way, I intend to share the results of the research freely with all those who participate. We intend to publish an initial analysis for research partners in September 2009. You can join the research by commenting here, or email me
Entry filed under: Uncategorized. Tags: .
1.
joeadamfry | 03/08/2009 at 11:45 am
Here you go: a stream of conciousness
VUI is awesome -but I still think its not ready. The iPhone VC is average, the AI of the VUID is still quite weak, it seems…essentially the old problem of distinguishing commands from general diagetic sound remains problemtaic: old schoolers in Star Trek etc going “COMPUTER” as an initiating command is plausible but surreal.
The specificity of the command is also an interesting point: can you say “send a message to x”, “text x”, send x a text”, “tell x y” etc to mean text x? Or do you have to say a specific string?
Further, I think I just feel like a bit of a tool talking to my phone. The thing I’d most like to be able to do is shout for it when I’ve lost it and for it to reply by beeping or shouting “I’m under the sofa cushion”…that would be of value. Especially if it could use an AR camera thing to work out where it was and direct me using his compass! Jokes.
Dictating eg a WP doc is something I’d be interested in as a I do a lot of research writing up & transcription – if my mac could auto transcribe from an audio recording that would be sick…again I think there would be issues of multiple voice/clarity etc. Still quite a niche proposition I find at the moment though.
I’d like to be able to shout ‘PAUSE’, ‘GO’, TURN IT DOWN’ at my stereo/laptop, I’d liek to be able to wake up and say ‘COMPUTER, NEWS’ and it tell me news headlines….I’d like to be able to shout for my iPhone and it respond and let me find it. but I don’t really give about VUI for the time being. What about multiple devices, all controlled by VUI…”LIGHTS< DOWN< STEREO<ON<MARVIN GAYE<CANDLES< LIGHT<OVEN<MAKE RISOTTO <FRIDGE<CHILL WINE" … or would they get mixed up with one another? I like the idea of remote control wars (when 2 people want to watch two different channels) occurring with only voice commands. Argh. http://joeadamfry.com
2.
joeadamfry | 03/08/2009 at 11:52 am
And yea, re the anthropomorphic character of VUI, I think yea, it does make a device more like a person – if they have a name and a voice etc…but they’re not a person, are they? If you’re talking about driving loyalty etc through the personality of a device (I’m thinking of Kit) then surely this would need to be highly customisable etc? Eg I really dont strike any affinity with the woman’s voice on iPhone 3GS VUI…in fact, I want to smack her. Also I think this relationship would only develop if communication was 2-way eg initiated by VUI AI rather than simply subservient.