Live data from Hacker News

Google Treks

google.com

21–30 of 114 posts

Re: Google Treks

#21

I have heard Google plans to do street-view of Burning Man, but it has never appeared. Perhaps this will be the venue if/when it finally does.

Probably not going to happen. The organizers require (or used to require) a percentage of all revenue made from professionally taken photos [0]

[0] http://www.jwz.org/gruntle/burningman.html

Re: Google Treks

#22

Where do I apply and what is the hourly rate? I assume all travel expenses are also paid. On a serious note, in the footsteps of OpenStreetMap that could be a great next community project: write the software for the infrastructure and the camera gear, open the designs of data collecting hardware (for the 3D printers) and let people upload their treks.

Elphel has been working on open source (hardware, software, board designs, etc) photogrammetric cameras for a while:

http://www3.elphel.com/eyesis-4pi

Re: Google Treks

#23
post #21

I have heard Google plans to do street-view of Burning Man, but it has never appeared. Perhaps this will be the venue if/when it finally does.

Probably not going to happen. The organizers require (or used to require) a percentage of all revenue made from professionally taken photos [0] [0] http://www.jwz.org/gruntle/burningman.html

That's crap.

Edit: damn, we need more people like jwz.

Re: Google Treks

#24
post #3

Google can't wait to usher in a future where people stay in their rooms and eat Soylent while being entertained and exercised by their VR-enabled computers.

Arguably these are great advertisements for the places they represent. If you were choosing between places to travel to, I think this would be helpful.

Re: Google Treks

#25
post #13

Just because it's from Google it must be special? Don't want to sound too negative but wondering if Trek would get any attention if done by some startup.

I'm not sure you appreciate the engineering infrastructure at work here. This isn't special because Google is doing it, it's special because it's really hard to do. How would you go about attempting this project as effectively as Google with the resources of a typical startup company?

Crawl the web for images, use the GPS EXIF data to find clusters, and use image processing to stitch together 3D environments from the photos.

It would not be as good as streetview though.

Re: Google Treks

#26
post #3

Google can't wait to usher in a future where people stay in their rooms and eat Soylent while being entertained and exercised by their VR-enabled computers.

That's pretty short-sighted.

Imagine if we had these kind of views of, say, the streets of London in 1900? This is worth doing for history's sake alone.

Re: Google Treks

#27

Earlier quoted context omitted.

I'm not sure you appreciate the engineering infrastructure at work here. This isn't special because Google is doing it, it's special because it's really hard to do. How would you go about attempting this project as effectively as Google with the resources of a typical startup company?

Crawl the web for images, use the GPS EXIF data to find clusters, and use image processing to stitch together 3D environments from the photos. It would not be as good as streetview though.

Creative...but I don't think it would have the same fluidity to Google's Trek. The images wouldn't be as up to date or "real time" as they could be because you'd be caching images that are potentially several days or weeks old.

And if you scripted a time limit for images, e.g. only within a few hours, you'd heavily limit the amount of usable images you had.

With "agents in the field" Google could potentially update its feeds whenever and as often as it needed. The difference would be having an in-house image-generating infrastructure or...well...tag-searching Instagram.

Re: Google Treks

#28
I would do this in a heartbeat, even for a substantially reduced salary. Assuming travel expenses and meals/shelter are paid for (within reason) I'd travel all over the world for an opportunity like this. Awesome. Google's been coming under a lot of fire lately but between this and Google Loon[1] I think Google's cooking up some really innovative, practically useful and genuinely creative toys for the world to play with.

[1]: http://www.google.com/loon/#utm_source=google&utm_medium=cpc...

Re: Google Treks

#30

Earlier quoted context omitted.

I'm not sure you appreciate the engineering infrastructure at work here. This isn't special because Google is doing it, it's special because it's really hard to do. How would you go about attempting this project as effectively as Google with the resources of a typical startup company?

Crawl the web for images, use the GPS EXIF data to find clusters, and use image processing to stitch together 3D environments from the photos. It would not be as good as streetview though.

Microsoft did this, effectively. Someone who remembers the project can provide a link to the demo/talks (TED, if I recall) hopefully.

edit: believe it was related to http://en.wikipedia.org/wiki/Photosynth

Post reply on HN