mycosystems/midtowndrafting.com
4
0
Fork 0
Code Pull requests Activity Actions

fact check, 3d gaussian splatting should not be roped in with other neural network based methods

Browse source
This commit is contained in:
brooke 2024-06-03 03:51:40 -04:00
parent a29598ccbe
commit 3f0a90d725
1 changed files with 4 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

8
content/docs/3d-scanning/nerf/_index.md
View file

@ -1,5 +1,5 @@
---
title: NeRF
title: Radiance Fields
---
<svg style="width:60%; margin-top:2rem;" viewBox="2.9938 7.971 92.0062 32.0289" xmlns="http://www.w3.org/2000/svg">
@ -19,11 +19,11 @@ title: NeRF
## Intro
The field of 3D scanning encompasses a wide range of applications, [NeRF](https://www.matthewtancik.com/nerf) makes a few of those applications a lot easier. The concept of using 3D scanning as an alternative means for people to interact with their surroundings is intriguing. For instance, museums could 3D scan their entire collections, or capture high-quality scans of monuments at risk of being lost forever. Additionally, education can be enhanced by introducing more active learning methods based on working in 3D. All of this is better served by how easy it is to produce models using a *more* software-defined method like NeRF.
The concept of using 3D scanning as an alternative means for people to interact with their surroundings is intriguing. Museums could 3D scan their entire collections, capturing high-quality scans of artifacts and exhibits that can be shared with a global audience. This technology also enables the preservation of monuments at risk of being lost forever, allowing people to explore and learn from these cultural features remotely. Moreover, education can be enhanced by introducing more active learning methods based on working in 3D, which differing methods of constructing [radiance fields](https://www.reedbeta.com/blog/the-radiance-field/) make simpler.
Though techniques such as NeRF, [MIP-NeRF](https://arxiv.org/abs/2103.13415), [3D Gaussians](https://arxiv.org/abs/2308.04079), and others demonstrate a fast and simpler method of 3D scanning compared to traditional laser or projection scanning. Neural network based scanning is still not considered a perfect "production ready" method, these models are made up of very dense point clouds that can not easily be translated into CAD. Supplimentary software is getting better but laser and projection scanning is still the best for high-accuracy work.
Techniques such as NeRF, [MIP-NeRF](https://arxiv.org/abs/2103.13415), [3D Gaussian Splatting](https://arxiv.org/abs/2308.04079), and others demonstrate a fast and simpler method of 3D scanning compared to traditional laser or projection scanning. These methods are not considered "production-ready." These models are made up of very dense "point clouds" that cannot easily be translated into CAD. Supplementary software is getting better, but laser and projection scanning are still the best for high-accuracy work.
3D guassians have been a new focus of research, you can use an entirely software-defined approach to create very light (the below model is 6MB) and high quality 3D models. With [polygon-based](https://en.wikipedia.org/wiki/Polygonal_modeling) models you would have never been able to get this amount of accuracy in a 3D model that can run on a phone.
3D guassians have been a new focus of research. You can use an entirely software-defined approach to create very light (the below model is 6 MB) and high-quality 3D models. With [polygon-based](https://en.wikipedia.org/wiki/Polygonal_modeling) models, you would have never been able to get this amount of accuracy in a 3D model that can run on a phone.
## Showcase