# Depth-Guided Robust and Fast Point Cloud Fusion NeRF for Sparse Input Views

arXiv: https://arxiv.org/abs/2403.02063 · HTML: https://mlsanity.com/paper/2403.02063/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Builds on 15 of its 27 references that are in the map.
- First posted on arXiv 4 Mar 2024.

## Builds on (most cited first, 15 of 15)

- [PyTorch: An Imperative Style, High-Performance Deep Learning Library](https://mlsanity.com/paper/1912.01703/) (2019)
- [pixelNeRF: Neural Radiance Fields from One or Few Images](https://mlsanity.com/paper/2012.02190/) (2020)
- [Depth-supervised NeRF: Fewer Views and Faster Training for Free](https://mlsanity.com/paper/2107.02791/) (2021)
- [Local Light Field Fusion: Practical View Synthesis with Prescriptive Sampling Guidelines](https://mlsanity.com/paper/1905.00889/) (2019)
- [RegNeRF: Regularizing Neural Radiance Fields for View Synthesis from Sparse Inputs](https://mlsanity.com/paper/2112.00724/) (2021)
- [Putting NeRF on a Diet: Semantically Consistent Few-Shot View Synthesis](https://mlsanity.com/paper/2104.00677/) (2021)
- [DeepVoxels: Learning Persistent 3D Feature Embeddings](https://mlsanity.com/paper/1812.01024/) (2018)
- [FreeNeRF: Improving Few-shot Neural Rendering with Free Frequency Regularization](https://mlsanity.com/paper/2303.07418/) (2023)
- [Dense Depth Priors for Neural Radiance Fields from Sparse Input Views](https://mlsanity.com/paper/2112.03288/) (2021)
- [SPARF: Neural Radiance Fields from Sparse and Noisy Poses](https://mlsanity.com/paper/2211.11738/) (2022)
- [InfoNeRF: Ray Entropy Minimization for Few-Shot Neural Volume Rendering](https://mlsanity.com/paper/2112.15399/) (2021)
- [Streaming Radiance Fields for 3D Video Synthesis](https://mlsanity.com/paper/2210.14831/) (2022)
- [MeshNet: Mesh Neural Network for 3D Shape Representation](https://mlsanity.com/paper/1811.11424/) (2018)
- [Kinematic-Structure-Preserved Representation for Unsupervised 3D Human Pose Estimation](https://mlsanity.com/paper/2006.14107/) (2020)
- [Boosting Point Clouds Rendering via Radiance Mapping](https://mlsanity.com/paper/2210.15107/) (2022)

## Built on by (most cited first, 6 of 6)

- [TranSplat: Generalizable 3D Gaussian Splatting from Sparse Multi-View Images with Transformers](https://mlsanity.com/paper/2408.13770/) (2024)
- [4D Gaussian Splatting in the Wild with Uncertainty-Aware Regularization](https://mlsanity.com/paper/2411.08879/) (2024)
- [MarkSplatter: Generalizable Watermarking for 3D Gaussian Splatting Model via Splatter Image Structure](https://mlsanity.com/paper/2509.00757/) (2025)
- [BRDF-NeRF: Neural Radiance Fields with Optical Satellite Images and BRDF Modelling](https://mlsanity.com/paper/2409.12014/) (2024)
- [Simple-RF: Regularizing Sparse Input Radiance Fields with Simpler Solutions](https://mlsanity.com/paper/2404.19015/) (2024)
- [RDG-GS: Relative Depth Guidance with Gaussian Splatting for Real-time Sparse-View 3D Rendering](https://mlsanity.com/paper/2501.11102/) (2025)

## Abstract (arXiv)

Novel-view synthesis with sparse input views is important for real-world applications like AR/VR and autonomous driving. Recent methods have integrated depth information into NeRFs for sparse input synthesis, leveraging depth prior for geometric and spatial understanding. However, most existing works tend to overlook inaccuracies within depth maps and have low time efficiency. To address these issues, we propose a depth-guided robust and fast point cloud fusion NeRF for sparse inputs. We perceive radiance fields as an explicit voxel grid of features. A point cloud is constructed for each input view, characterized within the voxel grid using matrices and vectors. We accumulate the point cloud of each input view to construct the fused point cloud of the entire scene. Each voxel determines its density and appearance by referring to the point cloud of the entire scene. Through point cloud fusion and voxel grid fine-tuning, inaccuracies in depth values are refined or substituted by those from other views. Moreover, our method can achieve faster reconstruction and greater compactness through effective vector-matrix decomposition. Experimental results underline the superior performance and time efficiency of our approach compared to state-of-the-art baselines.