FaithFusion: Harmonizing Reconstruction and Generation via Pixel-wise Information Gain

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, YuAn, Li, Xiaofan, Huang, Chi, Zhang, Wenhao, Li, Hao, Wang, Bosheng, Sun, Xun, Wang, Jun
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912729959235584
author Wang, YuAn
Li, Xiaofan
Huang, Chi
Zhang, Wenhao
Li, Hao
Wang, Bosheng
Sun, Xun
Wang, Jun
author_facet Wang, YuAn
Li, Xiaofan
Huang, Chi
Zhang, Wenhao
Li, Hao
Wang, Bosheng
Sun, Xun
Wang, Jun
contents In controllable driving-scene reconstruction and 3D scene generation, maintaining geometric fidelity while synthesizing visually plausible appearance under large viewpoint shifts is crucial. However, effective fusion of geometry-based 3DGS and appearance-driven diffusion models faces inherent challenges, as the absence of pixel-wise, 3D-consistent editing criteria often leads to over-restoration and geometric drift. To address these issues, we introduce \textbf{FaithFusion}, a 3DGS-diffusion fusion framework driven by pixel-wise Expected Information Gain (EIG). EIG acts as a unified policy for coherent spatio-temporal synthesis: it guides diffusion as a spatial prior to refine high-uncertainty regions, while its pixel-level weighting distills the edits back into 3DGS. The resulting plug-and-play system is free from extra prior conditions and structural modifications.Extensive experiments on the Waymo dataset demonstrate that our approach attains SOTA performance across NTA-IoU, NTL-IoU, and FID, maintaining an FID of 107.47 even at 6 meters lane shift. Our code is available at https://github.com/wangyuanbiubiubiu/FaithFusion.
format Preprint
id arxiv_https___arxiv_org_abs_2511_21113
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle FaithFusion: Harmonizing Reconstruction and Generation via Pixel-wise Information Gain
Wang, YuAn
Li, Xiaofan
Huang, Chi
Zhang, Wenhao
Li, Hao
Wang, Bosheng
Sun, Xun
Wang, Jun
Computer Vision and Pattern Recognition
In controllable driving-scene reconstruction and 3D scene generation, maintaining geometric fidelity while synthesizing visually plausible appearance under large viewpoint shifts is crucial. However, effective fusion of geometry-based 3DGS and appearance-driven diffusion models faces inherent challenges, as the absence of pixel-wise, 3D-consistent editing criteria often leads to over-restoration and geometric drift. To address these issues, we introduce \textbf{FaithFusion}, a 3DGS-diffusion fusion framework driven by pixel-wise Expected Information Gain (EIG). EIG acts as a unified policy for coherent spatio-temporal synthesis: it guides diffusion as a spatial prior to refine high-uncertainty regions, while its pixel-level weighting distills the edits back into 3DGS. The resulting plug-and-play system is free from extra prior conditions and structural modifications.Extensive experiments on the Waymo dataset demonstrate that our approach attains SOTA performance across NTA-IoU, NTL-IoU, and FID, maintaining an FID of 107.47 even at 6 meters lane shift. Our code is available at https://github.com/wangyuanbiubiubiu/FaithFusion.
title FaithFusion: Harmonizing Reconstruction and Generation via Pixel-wise Information Gain
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2511.21113