# KeyGS: A Keyframe-Centric Gaussian Splatting Method for Monocular Image Sequences

arXiv: https://arxiv.org/abs/2412.20767 · HTML: https://mlsanity.com/paper/2412.20767/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 15 of its 22 references that are in the map.
- First posted on arXiv 30 Dec 2024.

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

- [3D Gaussian Splatting for Real-Time Radiance Field Rendering](https://mlsanity.com/paper/2308.04079/) (2023)
- [Vision Transformers for Dense Prediction](https://mlsanity.com/paper/2103.13413/) (2021)
- [Direct Voxel Grid Optimization: Super-fast Convergence for Radiance Fields Reconstruction](https://mlsanity.com/paper/2111.11215/) (2021)
- [PlenOctrees for Real-time Rendering of Neural Radiance Fields](https://mlsanity.com/paper/2103.14024/) (2021)
- [BARF: Bundle-Adjusting Neural Radiance Fields](https://mlsanity.com/paper/2104.06405/) (2021)
- [NICE-SLAM: Neural Implicit Scalable Encoding for SLAM](https://mlsanity.com/paper/2112.12130/) (2021)
- [iMAP: Implicit Mapping and Positioning in Real-Time](https://mlsanity.com/paper/2103.12352/) (2021)
- [PhysGaussian: Physics-Integrated 3D Gaussians for Generative Dynamics](https://mlsanity.com/paper/2311.12198/) (2023)
- [A Hierarchical 3D Gaussian Representation for Real-Time Rendering of Very Large Datasets](https://mlsanity.com/paper/2406.12080/) (2024)
- [Self-Calibrating Neural Radiance Fields](https://mlsanity.com/paper/2108.13826/) (2021)
- [Progressively Optimized Local Radiance Fields for Robust View Synthesis](https://mlsanity.com/paper/2303.13791/) (2023)
- [Parallel Inversion of Neural Radiance Fields for Robust Pose Estimation](https://mlsanity.com/paper/2210.10108/) (2022)
- [Robust Camera Pose Refinement for Multi-Resolution Hash Encoding](https://mlsanity.com/paper/2302.01571/) (2023)
- [Improving Robustness for Joint Optimization of Camera Poses and Decomposed Low-Rank Tensorial Radiance Fields](https://mlsanity.com/paper/2402.13252/) (2024)
- [BAA-NGP: Bundle-Adjusting Accelerated Neural Graphics Primitives](https://mlsanity.com/paper/2306.04166/) (2023)

## Abstract (arXiv)

Reconstructing high-quality 3D models from sparse 2D images has garnered significant attention in computer vision. Recently, 3D Gaussian Splatting (3DGS) has gained prominence due to its explicit representation with efficient training speed and real-time rendering capabilities. However, existing methods still heavily depend on accurate camera poses for reconstruction. Although some recent approaches attempt to train 3DGS models without the Structure-from-Motion (SfM) preprocessing from monocular video datasets, these methods suffer from prolonged training times, making them impractical for many applications. In this paper, we present an efficient framework that operates without any depth or matching model. Our approach initially uses SfM to quickly obtain rough camera poses within seconds, and then refines these poses by leveraging the dense representation in 3DGS. This framework effectively addresses the issue of long training times. Additionally, we integrate the densification process with joint refinement and propose a coarse-to-fine frequency-aware densification to reconstruct different levels of details. This approach prevents camera pose estimation from being trapped in local minima or drifting due to high-frequency signals. Our method significantly reduces training time from hours to minutes while achieving more accurate novel view synthesis and camera pose estimation compared to previous methods.