# AnchorSplat: Fast and Structure Consistent Detail Synthesis for Gaussian Splatting

arXiv: https://arxiv.org/abs/2607.01290 · HTML: https://mlsanity.com/paper/2607.01290/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 34 of its 47 references that are in the map.
- First posted on arXiv 1 Jul 2026.

## Builds on (most cited first, 30 of 34)

- [3D Gaussian Splatting for Real-Time Radiance Field Rendering](https://mlsanity.com/paper/2308.04079/) (2023)
- [PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation](https://mlsanity.com/paper/1612.00593/) (2016)
- [PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space](https://mlsanity.com/paper/1706.02413/) (2017)
- [SwinIR: Image Restoration Using Swin Transformer](https://mlsanity.com/paper/2108.10257/) (2021)
- [Image Super-Resolution via Iterative Refinement](https://mlsanity.com/paper/2104.07636/) (2021)
- [Point Transformer](https://mlsanity.com/paper/2011.00931/) (2020)
- [Image Super-Resolution Using Deep Convolutional Networks](https://mlsanity.com/paper/1501.00092/) (2014)
- [4D Gaussian Splatting for Real-Time Dynamic Scene Rendering](https://mlsanity.com/paper/2310.08528/) (2023)
- [2D Gaussian Splatting for Geometrically Accurate Radiance Fields](https://mlsanity.com/paper/2403.17888/) (2024)
- [Point Transformer V3: Simpler, Faster, Stronger](https://mlsanity.com/paper/2312.10035/) (2023)
- [Point Transformer V2: Grouped Vector Attention and Partition-based Pooling](https://mlsanity.com/paper/2210.05666/) (2022)
- [Deeply-Recursive Convolutional Network for Image Super-Resolution](https://mlsanity.com/paper/1511.04491/) (2015)
- [NeRF-SR: High-Quality Neural Radiance Fields using Supersampling](https://mlsanity.com/paper/2112.01759/) (2021)
- [Large Spatial Model: End-to-end Unposed Images to Semantic 3D](https://mlsanity.com/paper/2410.18956/) (2024)
- [Mamba3D: Enhancing Local Features for 3D Point Cloud Analysis via State Space Model](https://mlsanity.com/paper/2404.14966/) (2024)
- [DiffGS: Functional Gaussian Splatting Diffusion](https://mlsanity.com/paper/2410.19657/) (2024)
- [VideoGigaGAN: Towards Detail-rich Video Super-Resolution](https://mlsanity.com/paper/2404.12388/) (2024)
- [SRGS: Super-Resolution 3D Gaussian Splatting](https://mlsanity.com/paper/2404.10318/) (2024)
- [Advancements in Point Cloud Data Augmentation for Deep Learning: A Survey](https://mlsanity.com/paper/2308.12113/) (2023)
- [S2Gaussian: Sparse-View Super-Resolution 3D Gaussian Splatting](https://mlsanity.com/paper/2503.04314/) (2025)
- [DiSR-NeRF: Diffusion-Guided View-Consistent Super-Resolution NeRF](https://mlsanity.com/paper/2404.00874/) (2024)
- [SuperGaussian: Repurposing Video Models for 3D Super Resolution](https://mlsanity.com/paper/2406.00609/) (2024)
- [3DSceneEditor: Controllable 3D Scene Editing with Gaussian Splatting](https://mlsanity.com/paper/2412.01583/) (2024)
- [Super-NeRF: View-consistent Detail Generation for NeRF super-resolution](https://mlsanity.com/paper/2304.13518/) (2023)
- [GaussianSR: 3D Gaussian Super-Resolution with 2D Diffusion Priors](https://mlsanity.com/paper/2406.10111/) (2024)
- [UltraVSR: Achieving Ultra-Realistic Video Super-Resolution with Efficient One-Step Diffusion Space](https://mlsanity.com/paper/2505.19958/) (2025)
- [Current World Models Lack a Persistent State Core](https://mlsanity.com/paper/2606.20545/) (2026)
- [Infrared Image Super-Resolution: Systematic Review, and Future Trends](https://mlsanity.com/paper/2212.12322/) (2022)
- [Advancing 3D Point Cloud Understanding through Deep Transfer Learning: A Comprehensive Survey](https://mlsanity.com/paper/2407.17877/) (2024)
- [GAP: Gaussianize Any Point Clouds with Text Guidance](https://mlsanity.com/paper/2508.05631/) (2025)

## Abstract (arXiv)

3D Gaussian Splatting (3DGS) has emerged as a powerful representation for high-fidelity rendering. However, existing assets often suffer from quality bottlenecks such as missing details and texture noise. Prior attempts to enhance these assets via 2D image processing introduce multi-view inconsistencies and high computational costs. In this paper, we propose a novel 3D-native refinement paradigm named AnchorSplat. AnchorSplat is an end-to-end deep network operating directly on 3D structures, avoiding the expensive optimization overhead of traditional 3D-2D-3D pipelines. Crucially, AnchorSplat is a strictly source-free solution requiring no original multi-view images. Central to the proposed method is the Point Anchor Mechanism, which enforces geometric consistency via local offset constraints, mitigating ill-posed mapping and gradient confounding. Furthermore, AnchorSplat replaces iterative densification with a single-pass multiplication mechanism. To facilitate research, we construct 3DGS-SR, the first large-scale benchmark for this task. Experiments demonstrate state-of-the-art results on the 3DGS-SR dataset, with throughput up to 105 times faster than optimization methods. Notably, AnchorSplat exhibits robust zero-shot generalization across diverse data distributions, including generative model outputs and real-world scans.