Hacker News new | ask | show | jobs
by websiteapi 162 days ago
I'm surprised no one has tried to skip cameras all together and use ultrasound. mics use two or maybe even three orders of magnitude for an audio -> object inference stack vs visual. of course you can't detect colors or do A LOT of things, but hey, you could make glasses really look like regular glasses, especially if you got rid of the screens too.
4 comments

>you could make glasses really look like regular glasses

The cameras are not what makes the glasses bulky and people find a lot of utility in taking and sharing pictures and videos from their glasses. So you'll probably always want to have at least one camera on the product for that use case.

I'm not talking about bulk - I'm talking about the fact that regular glasses don't have cameras on them, thus don't look like regular glasses.
Glasses as a computer form factor is not really proven out yet, but cameras on the glasses are one of the things that people are actually using the Meta Raybans for. One of the primary things people do with them is capture POV video. Take away the cameras and you're left with what. ChatGPT on command and headphones and that's it? The Humane Pin would like a word. People buy smart glasses specifically for a rich feature set, the more the better (because it's a nerd/early adopter product as of now).

And also in the real world people just do not care about cameras on glasses as much as people on HN trot out the glasshole articles from a decade ago. Both smart glasses and phones that are actively recording are everywhere already.

Well yeah because the rayban has a camera that's what people buy it for. It hardly does anything else (at least the one without screen).

I'd explicitly want one without camera to avoid the 'glasshole effect'.

And yes people do care at least here in Europe. The meta glasses are banned at a lot of events now.

I've yet to see someone wear meta ray bans at work, so at least for me DoA. you're right tho. as for use case, more so a better type of Siri. presumably mics would make it much cheaper as well, on the order of a regular watch (~$100)
> I'm surprised no one has tried to skip cameras all together and use ultrasound.

The ratios of image resolution and viewing distance to physical size are veeeeeeery bad with sound compared to cameras though. Cameras are also completely passive sensors that don't require an attached emitter in most circumstances.

They tried it back with the Powerglove.

Not sure why you think we have off the shelf miniaturized sonar hardware at scale and shape detection tech that could beat out mobile cameras and computer vision software.

Pretty sure Wayne Tech had the prototype of this sonar-vision translation layer of software all wrapped up way back in 2008, so it’s just a matter of productizing that, and since Pickle seems to deal in fiction already there’s good product synergy.
> Not sure why you think we have off the shelf miniaturized sonar hardware at scale and shape detection tech that could beat out mobile cameras and computer vision software.

uhm I didn't say that - what I am asking is exactly the opposite, in fact and the power glove thing was hardly capitalized. I wouldn't consider that a serious attempt.