SeasonScapes: Learning Large-scale Re-lightable 3D Landscapes with Seasonal Variation from Sparse Webcams

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Kleger, Timo, Ma, Qi, Zhang, Deheng, Van Gool, Luc, Paudel, Danda Pani
Format: Preprint
Veröffentlicht: 2026
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866910205445406720
author Kleger, Timo
Ma, Qi
Zhang, Deheng
Van Gool, Luc
Paudel, Danda Pani
author_facet Kleger, Timo
Ma, Qi
Zhang, Deheng
Van Gool, Luc
Paudel, Danda Pani
contents We introduce SeasonScapes framework and a the SeasonScapes dataset: Swiss Sparse-view Mountain Scenes with Seasonal Changes that covers over 50 km x 60 km, composed of more than 85,000 webcam images captured from 32 different locations across 13 timestamps throughout a full year. By projecting these timestamp-specific images onto a 3D mesh, we construct seasonal 3D landscapes that reflect natural appearance changes over time. To address occlusions and missing data, we leverage conditional diffusion models for image-guided inpainting directly on the mesh. The resulting completed meshes can be further relighted using standard physically-based renderer.
format Preprint
id arxiv_https___arxiv_org_abs_2605_09039
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle SeasonScapes: Learning Large-scale Re-lightable 3D Landscapes with Seasonal Variation from Sparse Webcams
Kleger, Timo
Ma, Qi
Zhang, Deheng
Van Gool, Luc
Paudel, Danda Pani
Computer Vision and Pattern Recognition
We introduce SeasonScapes framework and a the SeasonScapes dataset: Swiss Sparse-view Mountain Scenes with Seasonal Changes that covers over 50 km x 60 km, composed of more than 85,000 webcam images captured from 32 different locations across 13 timestamps throughout a full year. By projecting these timestamp-specific images onto a 3D mesh, we construct seasonal 3D landscapes that reflect natural appearance changes over time. To address occlusions and missing data, we leverage conditional diffusion models for image-guided inpainting directly on the mesh. The resulting completed meshes can be further relighted using standard physically-based renderer.
title SeasonScapes: Learning Large-scale Re-lightable 3D Landscapes with Seasonal Variation from Sparse Webcams
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2605.09039