OX-NeRF: 3D X-ray Tomography Reconstruction from Sparse Views Using Implicit Neural Representation
To address this problem we present a new framework, Optimised X-ray Neural Radiance Fields (OX-NeRF), that combines cross-scene feature learning with scene-specific optimisation to reconstruct sets of related scenes.
ProofPaper ↗
Key points
- NeRF and Gaussian splatting methods have been successfully applied on X-ray scenes where the views are too sparse for 3D reconstruction via classical methods.
- Ultra-sparse scenes with 10 or fewer views such as those with high-rate or low-dose acquisition still, however, present a significant challenge.
- OX-NeRF employs a convolutional neural network (CNN) to identify cross-scene features while maintaining scene-specific multi-resolution hash grids of spatial features.
- Benchmarking on parallel-beam and cone-beam X-ray datasets shows OX-NeRF provides significantly higher reconstruction accuracy on ultra-sparse scenes compared to existing radiance field methods.
Sources (1)
- [1]OX-NeRF: 3D X-ray Tomography Reconstruction from Sparse Views Using Implicit Neural RepresentationarXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 8, 09:12 AM
To address this problem we present a new framework, Optimised X-ray Neural Radiance Fields (OX-NeRF), that combines cross-scene feature learning with scene-specific optimisation to reconstruct sets of related scenes.
NeRF and Gaussian splatting methods have been successfully applied on X-ray scenes where the views are too sparse for 3D reconstruction via classical methods.
Extractive summary: sentences quoted from the sources.