Gaussian splats 와 그 이후 후속 연구

gs

2025 년기준이긴한데 흐름을 알기위한 파생된 연구 빽업용

 

DepthRegularizedGS, SparseGS, DNGaussian.

Layout-guided Panoramic Gaussian Splatting For Indoor Roaming

Integrating Meshes and 3D Gaussians for Indoor Scene Reconstruction with SAM Mask Guidance

ODGS: 3D Scene Reconstruction from Omnidirectional Images with 3D Gaussian Splattings

OmniGS: Omnidirectional Gaussian Splatting for Fast Radiance Field Reconstruction using Omnidirectional Images

DreamScene360 : Unconstrained Text-to-3D Scene Generation with Panoramic Gaussian Splatting

 

1. image 2 3d (with multiview diffusion)
   • https://yuxuan-xue.com/human-3diffusion/
2. lighting 3개
   • https://arxiv.org/pdf/2406.09733
   • https://shunsukesaito.github.io/rgca/
   • https://nju-3dv.github.io/projects/Relightable3DGaussian/
3. face 특화
   • https://arxiv.org/pdf/2401.12900
   • https://shenhanqian.github.io/gaussian-avatars
   • https://arxiv.org/pdf/2401.12900
4. 전체 scene 배경
   • https://arxiv.org/pdf/2406.06521
   • https://realmdreamer.github.io/
   • https://zju3dv.github.io/pgsr/paper/pgsr.pdf
   • https://arxiv.org/pdf/2311.13384

 

TitleDescription

	diffusion query 에 style transfer 할 target 으로 쿼리하도록함 그와중에 부족한부부분 보정하는 알고리즘(cross domain style transfer)
A Construct-optimize Approach to Sparse View Synthesis Without Camera Pose	sparse view 로부터 coarse solution 을 만들었다함 depth map 이용
Portrait3D: Text-Guided High-Quality 3D Portrait Generation	3D diffusion baseline SDS에서 GAN 을 샘플링하여 학습시키고, 퀄리티를 올리기위해서 다시 diffusion model 에 넣어 랜더러를 optimize