WildPose: A Unified Framework for Robust Pose Estimation in the Wild

2026-05-12

AI Summary

This work addresses the limitations of existing monocular pose estimation methods in dynamic scenes, which often rely on static assumptions or are restricted to short sequences. The authors propose a unified pose estimation framework that, for the first time, integrates a frozen MASt3R pretrained backbone with a 3D-aware update operator, a multi-level motion mask detector, and differentiable bundle adjustment optimization. This approach achieves robust and accurate pose estimation in both static and dynamic environments without fine-tuning the feature extractor, thereby significantly enhancing generalization. The method consistently outperforms state-of-the-art approaches across multiple benchmarks, including Wild-SLAM and Bonn (dynamic), TUM and 7-Scenes (static), and Sintel (low ego-motion).

Abstract

Estimating camera pose in dynamic environments is a critical challenge, as most visual SLAM and SfM methods assume static scenes. While recent dynamic-aware methods exist, they are often not unified: semantic-based approaches are brittle, per-sequence optimization methods fail on short sequences, and other learned models may degrade on static-only scenes. We present WildPose, a unified monocular pose estimation framework that is robust in dynamic environments while maintaining state-of-the-art performance on static and low-ego-motion datasets. Our key insight is to connect two powerful paradigms in modern 3D vision: the rich perceptual frontend of feedforward models and the end-to-end optimization of differentiable bundle adjustment (BA). We achieve this with a 3D-aware update operator built on a frozen, pre-trained MASt3R feature backbone, together with a high-capacity motion mask detector that uses multi-level 3D-aware features from the same backbone. Extensive experiments show WildPose consistently outperforms prior methods across dynamic (Wild-SLAM, Bonn), static (TUM, 7-Scenes), and low-ego-motion (Sintel) benchmarks.

Problem

• pose estimation
• dynamic environments
• visual SLAM
• structure-from-motion
• camera pose

Innovation

• unified pose estimation
• dynamic scene robustness
• differentiable bundle adjustment
• 3D-aware feature backbone
• motion mask detection