ParticleSfM: Exploiting Dense Point Trajectories for Localizing Moving Cameras in the Wild

TL;DR

ParticleSfM introduces a dense, trajectory-based structure-from-motion approach that improves camera pose estimation in dynamic scenes by leveraging optical flow and motion segmentation, outperforming existing methods.

Contribution

The paper presents a novel neural network architecture for processing irregular point trajectories and a robust method for estimating camera pose in dynamic environments using dense point trajectories.

Findings

Significantly more accurate camera trajectories on MPI Sintel dataset.

Outperforms state-of-the-art dense correspondence methods in dynamic scenes.

Maintains reasonable accuracy in fully static scenes.

Abstract

Estimating the pose of a moving camera from monocular video is a challenging problem, especially due to the presence of moving objects in dynamic environments, where the performance of existing camera pose estimation methods are susceptible to pixels that are not geometrically consistent. To tackle this challenge, we present a robust dense indirect structure-from-motion method for videos that is based on dense correspondence initialized from pairwise optical flow. Our key idea is to optimize long-range video correspondence as dense point trajectories and use it to learn robust estimation of motion segmentation. A novel neural network architecture is proposed for processing irregular point trajectory data. Camera poses are then estimated and optimized with global bundle adjustment over the portion of long-range point trajectories that are classified as static. Experiments on MPI Sintel dataset show that our system produces significantly more accurate camera trajectories compared to existing state-of-the-art methods. In addition, our method is able to retain reasonable accuracy of camera poses on fully static scenes, which consistently outperforms strong state-of-the-art dense correspondence based methods with end-to-end deep learning, demonstrating the potential of dense indirect methods based on optical flow and point trajectories. As the point trajectory representation is general, we further present results and comparisons on in-the-wild monocular videos with complex motion of dynamic objects. Code is available at https://github.com/bytedance/particle-sfm.