Stitch4D: Sparse Multi-Location 4D Urban Reconstruction via Spatio-Temporal Interpolation
Hina Kogure, Kei Katsumata, Taiki Miyanishi, Komei Sugiura

TL;DR
Stitch4D is a novel 4D urban reconstruction framework that synthesizes intermediate views to handle sparse multi-location data, enabling coherent geometry and scene dynamics reconstruction in urban environments.
Contribution
It introduces a unified approach that explicitly compensates for missing spatial coverage by synthesizing bridge views and jointly optimizing observations, addressing a key challenge in sparse multi-location 4D reconstruction.
Findings
Stitch4D outperforms existing 4D reconstruction methods on the U-S4D benchmark.
Synthesizing intermediate views improves spatial coverage and reconstruction stability.
The approach achieves coherent geometry and smooth scene dynamics in sparse urban settings.
Abstract
Dynamic urban environments are often captured by cameras placed at spatially separated locations with little or no view overlap. However, most existing 4D reconstruction methods assume densely overlapping views. When applied to such sparse observations, these methods fail to reconstruct intermediate regions and often introduce temporal artifacts. To address this practical yet underexplored sparse multi-location setting, we propose Stitch4D, a unified 4D reconstruction framework that explicitly compensates for missing spatial coverage in sparse observations. Stitch4D (i) synthesizes intermediate bridge views to densify spatial constraints and improve spatial coverage, and (ii) jointly optimizes real and synthesized observations within a unified coordinate frame under explicit inter-location consistency constraints. By restoring intermediate coverage before optimization, Stitch4D prevents…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
