Diffusion-guided Generalizable Enhancer for Urban Scene Reconstruction

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Che, Henry, Wang, Jingkang, Chen, Yun, Yang, Ze, Manivasagam, Sivabalan, Urtasun, Raquel
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916036323835904
author Che, Henry
Wang, Jingkang
Chen, Yun
Yang, Ze
Manivasagam, Sivabalan
Urtasun, Raquel
author_facet Che, Henry
Wang, Jingkang
Chen, Yun
Yang, Ze
Manivasagam, Sivabalan
Urtasun, Raquel
contents Urban scene reconstruction from real-world observations has emerged as a powerful tool for self-driving development and testing. While current neural rendering approaches achieve high-fidelity rendering along the recorded trajectories, their quality degrades significantly under large viewpoint shifts, limiting the applicability for closed-loop simulation. Recent works have shown promising results in using diffusion models to enhance quality at these challenging viewpoints and distill improvements back into 3D representations. However, they often require costly per-scene optimization, and the distilled representations remain fragile and fail to generalize beyond limited synthesized views. To address these limitations, we propose GenRe, a novel diffusion-guided generalizable enhancer for urban scene reconstruction. GenRe takes as input any pretrained 3D Gaussian representation and fixes the deficiencies within a few minutes. By learning to distill generative priors across diverse scenes, GenRe produces robust and high-fidelity representation efficiently that generalizes reliably to challenging unseen viewpoints (e.g., lane change). Experiments show that GenRe outperforms existing methods in both quality and efficiency and benefits various downstream tasks, enabling robust and scalable sensor simulation for autonomous driving.
format Preprint
id arxiv_https___arxiv_org_abs_2605_22420
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Diffusion-guided Generalizable Enhancer for Urban Scene Reconstruction
Che, Henry
Wang, Jingkang
Chen, Yun
Yang, Ze
Manivasagam, Sivabalan
Urtasun, Raquel
Computer Vision and Pattern Recognition
Artificial Intelligence
Robotics
Urban scene reconstruction from real-world observations has emerged as a powerful tool for self-driving development and testing. While current neural rendering approaches achieve high-fidelity rendering along the recorded trajectories, their quality degrades significantly under large viewpoint shifts, limiting the applicability for closed-loop simulation. Recent works have shown promising results in using diffusion models to enhance quality at these challenging viewpoints and distill improvements back into 3D representations. However, they often require costly per-scene optimization, and the distilled representations remain fragile and fail to generalize beyond limited synthesized views. To address these limitations, we propose GenRe, a novel diffusion-guided generalizable enhancer for urban scene reconstruction. GenRe takes as input any pretrained 3D Gaussian representation and fixes the deficiencies within a few minutes. By learning to distill generative priors across diverse scenes, GenRe produces robust and high-fidelity representation efficiently that generalizes reliably to challenging unseen viewpoints (e.g., lane change). Experiments show that GenRe outperforms existing methods in both quality and efficiency and benefits various downstream tasks, enabling robust and scalable sensor simulation for autonomous driving.
title Diffusion-guided Generalizable Enhancer for Urban Scene Reconstruction
topic Computer Vision and Pattern Recognition
Artificial Intelligence
Robotics
url https://arxiv.org/abs/2605.22420