DROID-SLAM in the Wild

TL;DR

DROID-SLAM in the Wild is a real-time RGB SLAM system that robustly handles dynamic, cluttered environments by estimating per-pixel uncertainty from multi-view features, achieving state-of-the-art results in challenging scenarios.

Contribution

The paper introduces a novel dynamic SLAM approach that leverages differentiable uncertainty-aware bundle adjustment and multi-view feature inconsistency for robust tracking in real-world scenes.

Findings

Achieves state-of-the-art camera pose accuracy in dynamic environments.

Operates in real-time at around 10 FPS.

Effectively handles unknown dynamic objects and cluttered scenes.

Abstract

We present a robust, real-time RGB SLAM system that handles dynamic environments by leveraging differentiable Uncertainty-aware Bundle Adjustment. Traditional SLAM methods typically assume static scenes, leading to tracking failures in the presence of motion. Recent dynamic SLAM approaches attempt to address this challenge using predefined dynamic priors or uncertainty-aware mapping, but they remain limited when confronted with unknown dynamic objects or highly cluttered scenes where geometric mapping becomes unreliable. In contrast, our method estimates per-pixel uncertainty by exploiting multi-view visual feature inconsistency, enabling robust tracking and reconstruction even in real-world environments. The proposed system achieves state-of-the-art camera poses and scene geometry in cluttered dynamic scenarios while running in real time at around 10 FPS. Code and datasets are available at https://github.com/MoyangLi00/DROID-W.git.