MambaVO: Deep Visual Odometry Based on Sequential Matching Refinement and Training Smoothing

TL;DR

MambaVO introduces a robust deep visual odometry method that refines frame matching and training stability, achieving state-of-the-art accuracy and real-time performance in challenging scenarios.

Contribution

The paper presents a novel sequential matching refinement and training smoothing approach, improving pose estimation robustness in deep visual odometry.

Findings

Achieves state-of-the-art results on public benchmarks.

Ensures real-time performance in challenging environments.

Improves matching accuracy and training stability.

Abstract

Deep visual odometry has demonstrated great advancements by learning-to-optimize technology. This approach heavily relies on the visual matching across frames. However, ambiguous matching in challenging scenarios leads to significant errors in geometric modeling and bundle adjustment optimization, which undermines the accuracy and robustness of pose estimation. To address this challenge, this paper proposes MambaVO, which conducts robust initialization, Mamba-based sequential matching refinement, and smoothed training to enhance the matching quality and improve the pose estimation. Specifically, the new frame is matched with the closest keyframe in the maintained Point-Frame Graph (PFG) via the semi-dense based Geometric Initialization Module (GIM). Then the initialized PFG is processed by a proposed Geometric Mamba Module (GMM), which exploits the matching features to refine the overall inter-frame matching. The refined PFG is finally processed by differentiable BA to optimize the poses and the map. To deal with the gradient variance, a Trending-Aware Penalty (TAP) is proposed to smooth training and enhance convergence and stability. A loop closure module is finally applied to enable MambaVO++. On public benchmarks, MambaVO and MambaVO++ demonstrate SOTA performance, while ensuring real-time running.