Geo4D: leveraging video generators for geometric 4Dscene reconstruction

Conference item

We introduce Geo4D, a method to repurpose video diffusion models for monocular 3D reconstruction of dynamic scenes. By leveraging the strong dynamic priors captured by largescale pre-trained video models, Geo4D can be trained using only synthetic data while generalizing well to real data in a zero-shot manner. Geo4D predicts several complementary geometric modalities, namely point, disparity, and ray maps. We propose a new multi-modal alignment algorithm to align and fuse these modalities, as well as a sliding window approach at inference time, thus enabling robust and accurate 4D reconstruction of long videos. Extensive experiments across multiple benchmarks show that Geo4D significantly surpasses state-of-the-art video depth estimation methods.

Authors: Jiang, Z; Zheng, C; Laina, I; Larlus, D; Vedaldi, A

• Publication status: Accepted
• Peer review status: Peer reviewed
• Publisher: IEEE
• Acceptance date: 2025-07-23
• Event title: International Conference on Computer Vision (ICCV 2025)
• Event location: Honolulu, Hawai'i, USA
• Event website: https://iccv.thecvf.com/
• Event start date: 2025-10-19
• Event end date: 2025-10-23
• Language: English