marcusandreassvensson / gaussian-splatting-webgpu Goto Github PK
View Code? Open in Web Editor NEWHome Page: https://splatting.marcussvensson.com
License: MIT License
Home Page: https://splatting.marcussvensson.com
License: MIT License
Hello,
Thanks for your code.
First, we have started the project on my server.
Next, we have used the --host option to expose the network.
Then, we visit the website, and we can see the interface of the website.
After, we upload the ply file. The website can not show the 3D model. Moreover, it shows a message:
"Error: WebGPU not supported on this browser! (navigator.gpu is null)"
How can we solve this problem?
Today the sorting array size is set to 14mil entries to accommodate for larger scenes, this makes the performance suffer for smaller scenes more than necessary, but the amount of total splat-tile intersections is variable every viewpoint, which means we can't use a static array size based on the scene.
Hello, i tried using the Demo (https://splatting.marcussvensson.com/) with the example data any custom data on a rtx 3070 but the results were always completely black or consited of black horizontal lines. Is there a solutions to this? Great Project!
The last subpass in one sort pass just before the scatter accesses the shared memory in a strided way which means there could be a potential performance gain in the memory access pattern.
The rasterization shader has many opportunities for improvement.
The largest performance hit right now is the loop which handles the color contribution from every splat in the tile.
Some great optimizations would involve effectively culling and early stopping of the loop, less global memory fetches and less branching.
Compare to https://github.com/antimatter15/splat?tab=readme-ov-file webgl impl.
In my understanding, webGPU performance should be better. I guess because radixsort implemented on GPU, or do I miss something?
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.