It's free on YouTube: https://www.youtube.com/watch?v=3-Jrp6it9Ss
For example, check the Vienna map .. so many interesting locations!
If you want GS news, https://radiancefields.com/ reports a lot of advances all the time.
This captured that serenity.
There's a town hall on 5th March with speakers from Niantic and Cesium: https://metaverse-standards.org/event/gaussian-splats-town-h....
The previous splats town hall, and other related talks, are on the videos page (there was another gaussian splat talk a couple of days ago from Adobe). https://metaverse-standards.org/presentations-videos/
I tried it on my MQ3 last night and it was the first thing like that which was photorealistic, but it badly overloaded the MQ3, so it was the closest experience to Sword Art Online I've had yet in VR. (The sky was transparent and my room showed through!) I should have been sitting when I started it but since the frame rate was low and the horizon improperly oriented I could have fallen transitioning to the couch if I hadn't steeled myself to rely 100% on proprioception.
Contrasted to the way too lo-fi Inside the Scaniverse and the bland but cringe Horizon Worlds it's a hit. I gotta try it again in tethered mode.
The basic technique for rendering gaussian splats is kryptonite for this architecture, essentially implementing every worse practice for rendering on a mobile GPU:
* Tons of overdraw (overlapping splats)
* Tons of alpha blending
* Millions of splats in the distance generate a lot of tiny triangles resolving to a single pixel
* Long thin splats in the foreground generate triangles that cover multiple tiles
These are all the ingredients you need to bring a mobile GPU to its knees! Any desktop GPU (including most laptops) will be far less sensitive to these issues, even if it's not very powerful. It's a fundamental issue of architecture rather than one of raw FLOPs.
I had to power cycle mine to get out, but boy was the view great despite the motion sickness.
TIL — very cool work!
I've been trying to build something myself for biological and heritage captures. Could you elaborate at all on the JS decoder method used or resources that ended up enabling you to make it?
I've been playing with Playcanvas React for the past week, would your method integrate with their platform?
For more information on the format itself you can look at https://github.com/nerfstudio-project/gsplat/blob/main/gspla....
It's very similar to PlayCanvas React, but with a fully Web Components-based architecture. You can use the same scripts interchangeably between both frameworks.
I'm trying to make something that can take in real-time data and also handle some XR/AR if possible as well, but I am pretty ignorant beyond html/css/js.
Then you can make video games and interactive experiences in real-world locations. I doubt the collision handling is there, but you can at least start with something like Microsoft Flight Simulator with low-flying drones.
It would also be great for training AI to generate realistic-looking scenarios. Besides playing and working (ex: VR) in real places, you can in very-realistic liminal spaces. (And it's training on public areas, so less ethical issues.)
I have an extreme interest in deriving measurements from splat data of vegetation, as it tends to reconstruct thin planes like leaves far better than other traditional SfM techniques.
Something I’ve been looking for for a while is an interactive view of SF from above - I think it’d be cool to experience the verticality of SF and see how all the different hills relate to each other (and gmaps/earth just isn't cutting it).
this is actually pretty good for around the panhandle, but if anyone is aware of something like that for the whole city please let me know!
You can gate it with pointer capture API but please do it!!!!
The background looks tantalising but maybe a little more testing is needed... (e.g. for the most common browser/OS in the world).
I tried again today and it's still really slow but I do get the controls this time so I was able to use it. Very neat!
This is on a Pixel 8, Chrome 113.
I used to live nearby and my favorite “urban” hike was going up Glen Canyon, up and down the two Twin Peaks, loop through Sutro Forest and then go back to downtown walking down 17th. Loved those weekend walks.
One bit of feedback: don't move the camera if someone clicks one of the circles, that is super disorienting. There is also a bug that if a drag to move the camera happens to end on a circle, the popup opens.
Naively, it seems to me that the many needle-like artifacts further away from the tower could be filtered out?
Scanning the spectrum to pull in KQED, etc. The first Austin City Limits I saw in HD blew my mind.
I think all but one TV station came to the South Bay by way of Sutro. Quite a reach.
- https://maps.google.com (satellite view)
- https://earth.google.com (also in browser, possibly better camera controls for what you want)
- Bing Maps (3D flyover mode, more stale data in my experience)
- Apple Maps satellite view (only on macOS/iOS)
- Google Earth VR[1] (requires a Windows PC and a VR headset that can connect to it)
- Microsoft Flight Sim 2020/2024 (requires a beefy Windows PC, uses Bing Maps plus a lot of other enhancements and rendering goodness. Most lifelike "feels like I'm there" but not true to earth)
I'm not aware of splat-based city photogrammetry aside from one-offs like this but I'd love to learn if there's any such projects!
[1] https://store.steampowered.com/app/348250/Google_Earth_VR/
Touch controls are very weird and don't rotate the view as expected.
Which means effectively zero over-the-air reception in parts of the Mission. Lesson learned during a Super Bowl party the 2010s.
The video: https://www.youtube.com/watch?v=A7-3CKulsCc
One of the most desirable places on earth to live and it's on a small peninsula. Yet it's a sea of single-family homes as far as the eye can see.
The distance between Sutro Tower and the "Downtown" SF is less than the distance between the Brooklyn Bridge and Central Park. But could you imagine if that space was filled with 2-3 story townhomes?
It's a real travesty.
I live in the that space (between tower and the city) and the local neighbourhood group (HANC) is ridiculously NIMBY. rezoning is happening but it's slow going...
I hope the US gets its act together and learns from exemplar infrastructure projects around the world.
Ever wondered if it’s desirable BECAUSE it’s not a dense urban jungle?
Mild winters, mild summers.
Not too much rain.
No serious threat from tornadoes or hurricanes.
That’s a very big draw and it wouldn’t go away by making more dense housing, even if the rest of the peninsula was developed like Manhattan.
Londons problem for example is that they tried to be clever and cheap at the same time in the 60s and now we're stuck with it.
Considering the advances in seismic technology made over the past fifty years, it is a shame that much faster upgrades to the real estate have not been encouraged.
If you're going to force-densify anything, why not actual low-per-capita population areas [0] and develop mass transit, so North America can have the successful China city-tier model [1] with spread-out opportunities instead of cramming everyone together in one place.
[0] https://en.wikipedia.org/wiki/File:California_population_map...
> There would be 3-4X the people, yet still the same amount of roads, services, and public utilities.
That's the point! Per capita, it should be cheaper to live in cities because infrastructure goes so much further. And if you are arguing for better mass transit, you will have to build many, many more miles if you also want to encourage people to sprawl.
Although I think the strongest case for allowing cities to get dense is it allows greenbelts and less dense areas closer to the city. You can build a big dense city UP and make it easier for people to get out and enjoy nature and farms and etc. Or you can build a city OUT and then it's just desolate city for hours around.
The argument in favour of density is that if you increase density, then you also decrease the average distance that people have to travel until they get somewhere interesting, like a job or a shop.
Vehicle-delivered utilities like garbage collection, package deliveries, and mass transit get more efficient, and the same goes for tunnel-delivered utilities like fiber internet and water.
San Francisco is economically one of the world's most impactful cities; it'd be good for all of us if there was more of it. You get all sorts of interesting multiplier effects when you put lots of a certain kind of person in one place.
- all the theater kids in one town: LA
- all the bankers: NYC, London
- all the computer people: SF
So we have a decent mass transit system but its not far from peak capacity, and most of what the government has been doing is hacking around that. Trammish busses, cross river rail etc.
So we need to attack the issue from the other side too. We have a weirdly non dense central region, largely due to single issue anti development voters, who dont want apartment buildings right where they should be (on top of mass transit hubs). Instead the inner suburbs are littered with 1950s character homes, battleaxed once for massive profit.
We can take significant load off of a system close to a decade from collapse by simply removing outdated zoning.
And the way the council here operates, utilities and road upgrades necessitated by development are borne by the property developer. So there's really near zero cost in approaching things this way. And they have also used priority approvals, where if a certain amount of floor space of a development is earmarked for light commercial, they can cut a few years off approval time. So theres absolutely no reason not to, as the big residential buildings grow, they grow their own services and utilities.
For certain applications the internet can never compete with "broadcast".
