Talk:Hardware

Is it really true that voice and video input are less researched? I'd almost say the opposite! Eyetap system, by Steve Mann, works basically with video input, and he's been using and researching them for a long time. Also, can anyone count the number of news telling us that voice is the next keyboard? It surely has been researched, and been tried to implement a lot.

I'd say that voice input is kind of a fail really, but it's just my opinion. --OjM 21:45, 20 August 2009 (UTC)

Re: voice input.

I've used voice input - I've had RSI on and off for a few years until I had hand surgery, and when it was bad I used Dragon Naturally speaking 9 at work. It works surprisingly well. It was able to write letters and documents from spoken sentences. It was also really cool to be able to say start - programs - mozilla - firefox and see firefox pop up on screen.

Upsides: usable for commands and emails/documents, web browsing (read the link aloud and it works). Recognizes naturally spoken sentences best. Good accuracy, even better with training. Definitely usable.

Downsides: You feel like a twat talking to a computer until you get used to it. Probably even harder in public. Technical language (acronyms, commands) need to be spelled out the first time - then it 'learns' them, No Linux version :-( Microphone needs to be good quality and accurately positioned. Can be affected by background noise (random words appear) High CPU required (core 2 preferred - might be too much for small-board cpu)

Overall, I was far more impressed than when I tried it 10 years ago. It's now usable. If I was in a quiet office on my own, I would use it a lot more. Unfortunately keyboards are still slightly better - and you don't interrupt the person next to you.

For wearables, a specialised voice command list and a bluetooth headset might be a really useful addition to a powerful wearable.

--Renoir 19:36, 22 August 2009 (UTC)