SRSplat: Feed-Forward Super-Resolution Gaussian Splatting from Sparse Multi-View Images
Proceedings of the AAAI Conference on Artificial Intelligence · 14.03.2026 · 40 (6) · 4950-4958
DOI: 10.1609/aaai.v40i6.42499 ↗Kurz gefasst
Fachartikel in der Zeitschrift Proceedings of the AAAI Conference on Artificial Intelligence, online erschienen am 14.03.2026. Verfasst von Xinyuan Hu und 8 weiteren Autor:innen. Die Arbeit wurde bisher 1-mal zitiert (Crossref).
Abstract (Originalsprache)
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 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 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 RGFE. To further refine predicted Gaussian primitives, we introduce 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.
Zitieren
APA
Hu, X., Shi, C., Yang, C., Chen, M., Ding, J., Wei, T., Wei, C., Yu, Z., & Tan, M. (2026). SRSplat: Feed-Forward Super-Resolution Gaussian Splatting from Sparse Multi-View Images. Proceedings of the AAAI Conference on Artificial Intelligence, 40(6), 4950-4958. https://doi.org/10.1609/aaai.v40i6.42499
Vancouver
Hu X, Shi C, Yang C, Chen M, Ding J, Wei T, et al. SRSplat: Feed-Forward Super-Resolution Gaussian Splatting from Sparse Multi-View Images. Proceedings of the AAAI Conference on Artificial Intelligence. 2026;40(6):4950-4958. doi:10.1609/aaai.v40i6.42499
BibTeX
@article{Hu2026,
author = {Hu, Xinyuan and Shi, Changyue and Yang, Chuxiao and Chen, Minghao and Ding, Jiajun and Wei, Tao and Wei, Chen and Yu, Zhou and Tan, Min},
title = {SRSplat: Feed-Forward Super-Resolution Gaussian Splatting from Sparse Multi-View Images},
journal = {Proceedings of the AAAI Conference on Artificial Intelligence},
year = {2026},
volume = {40},
number = {6},
pages = {4950--4958},
doi = {10.1609/aaai.v40i6.42499}
}
EUVIMED