GeoMotion: Rethinking Motion Segmentation via Latent 4D Geometry

TL;DR

GeoMotion introduces a fully learning-based, end-to-end approach for motion segmentation that leverages latent 4D geometry and attention mechanisms, outperforming traditional methods in accuracy and efficiency.

Contribution

It proposes a novel deep learning framework that bypasses explicit correspondence estimation, utilizing latent features and 4D scene priors for robust motion segmentation.

Findings

Achieves state-of-the-art performance on motion segmentation benchmarks.

Eliminates the need for complex pre-processing and iterative refinement.

Demonstrates high efficiency and robustness in dynamic scene analysis.

Abstract

Motion segmentation in dynamic scenes is highly challenging, as conventional methods heavily rely on estimating camera poses and point correspondences from inherently noisy motion cues. Existing statistical inference or iterative optimization techniques that struggle to mitigate the cumulative errors in multi-stage pipelines often lead to limited performance or high computational cost. In contrast, we propose a fully learning-based approach that directly infers moving objects from latent feature representations via attention mechanisms, thus enabling end-to-end feed-forward motion segmentation. Our key insight is to bypass explicit correspondence estimation and instead let the model learn to implicitly disentangle object and camera motion. Supported by recent advances in 4D scene geometry reconstruction (e.g., $\pi^3$), the proposed method leverages reliable camera poses and rich spatial-temporal priors, which ensure stable training and robust inference for the model. Extensive experiments demonstrate that by eliminating complex pre-processing and iterative refinement, our approach achieves state-of-the-art motion segmentation performance with high efficiency. The code is available at:https://github.com/zjutcvg/GeoMotion.