SRSplat: Feed-Forward Super-Resolution Gaussian Splatting from Sparse Multi-View Images

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hu, Xinyuan, Shi, Changyue, Yang, Chuxiao, Chen, Minghao, Ding, Jiajun, Wei, Tao, Wei, Chen, Yu, Zhou, Tan, Min
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917081431146496
author Hu, Xinyuan
Shi, Changyue
Yang, Chuxiao
Chen, Minghao
Ding, Jiajun
Wei, Tao
Wei, Chen
Yu, Zhou
Tan, Min
author_facet Hu, Xinyuan
Shi, Changyue
Yang, Chuxiao
Chen, Minghao
Ding, Jiajun
Wei, Tao
Wei, Chen
Yu, Zhou
Tan, Min
contents Feed-forward 3D reconstruction from sparse, low-resolution (LR) images is a crucial capability for real-world applications, such as autonomous driving and embodied AI. However, existing methods often fail to recover fine texture details. This limitation stems from the inherent lack of high-frequency information in LR inputs. To address this, we propose \textbf{SRSplat}, a feed-forward framework that reconstructs high-resolution 3D scenes from only a few LR views. Our main insight is to compensate for the deficiency of texture information by jointly leveraging external high-quality reference images and internal texture cues. We first construct a scene-specific reference gallery, generated for each scene using Multimodal Large Language Models (MLLMs) and diffusion models. To integrate this external information, we introduce the \textit{Reference-Guided Feature Enhancement (RGFE)} module, which aligns and fuses features from the LR input images and their reference twin image. Subsequently, we train a decoder to predict the Gaussian primitives using the multi-view fused feature obtained from \textit{RGFE}. To further refine predicted Gaussian primitives, we introduce \textit{Texture-Aware Density Control (TADC)}, which adaptively adjusts Gaussian density based on the internal texture richness of the LR inputs. Extensive experiments demonstrate that our SRSplat outperforms existing methods on various datasets, including RealEstate10K, ACID, and DTU, and exhibits strong cross-dataset and cross-resolution generalization capabilities.
format Preprint
id arxiv_https___arxiv_org_abs_2511_12040
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SRSplat: Feed-Forward Super-Resolution Gaussian Splatting from Sparse Multi-View Images
Hu, Xinyuan
Shi, Changyue
Yang, Chuxiao
Chen, Minghao
Ding, Jiajun
Wei, Tao
Wei, Chen
Yu, Zhou
Tan, Min
Computer Vision and Pattern Recognition
Feed-forward 3D reconstruction from sparse, low-resolution (LR) images is a crucial capability for real-world applications, such as autonomous driving and embodied AI. However, existing methods often fail to recover fine texture details. This limitation stems from the inherent lack of high-frequency information in LR inputs. To address this, we propose \textbf{SRSplat}, a feed-forward framework that reconstructs high-resolution 3D scenes from only a few LR views. Our main insight is to compensate for the deficiency of texture information by jointly leveraging external high-quality reference images and internal texture cues. We first construct a scene-specific reference gallery, generated for each scene using Multimodal Large Language Models (MLLMs) and diffusion models. To integrate this external information, we introduce the \textit{Reference-Guided Feature Enhancement (RGFE)} module, which aligns and fuses features from the LR input images and their reference twin image. Subsequently, we train a decoder to predict the Gaussian primitives using the multi-view fused feature obtained from \textit{RGFE}. To further refine predicted Gaussian primitives, we introduce \textit{Texture-Aware Density Control (TADC)}, which adaptively adjusts Gaussian density based on the internal texture richness of the LR inputs. Extensive experiments demonstrate that our SRSplat outperforms existing methods on various datasets, including RealEstate10K, ACID, and DTU, and exhibits strong cross-dataset and cross-resolution generalization capabilities.
title SRSplat: Feed-Forward Super-Resolution Gaussian Splatting from Sparse Multi-View Images
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.12040