用专家组合提升惯性里程计跨载体泛化能力
MosaicIMU: Composing Carrier Experts for Generalizable Neural Inertial Odometry
- 基于原型路由选择载体专属专家特征，实现条件化模型融合
- 在未见载体上降低40%平均轨迹误差，10秒内减少34%位置误差
- 适合需快速适配新设备的边缘部署场景，支持轻量增量学习
可靠的惯性里程计对各类载体在外部感知不可靠时至关重要。基于学习的方法通过捕捉局部运动先验来减少积分漂移，但通常局限于特定载体，难以在异构平台间泛化。本文提出MosaicIMU，一种面向可泛化神经惯性里程计的载体条件化混合专家（MoE）预训练-适应框架。该方法利用基于原型的路由机制组合载体特定专家特征，解码局部速度与不确定性约束，并与历史感知卡尔曼滤波器融合。针对未见域适应，冻结预训练主干模型，仅学习轻量级专家残差分支；针对边缘部署，进一步复用路由机制筛选有效在线样本以实现高效增量更新。实验表明，MosaicIMU持续优于现有学习基线，平均绝对轨迹误差（ATE）和10秒内相对轨迹误差（RTE-10s）分别降低40%和34%。结果表明，MosaicIMU为可泛化、自适应神经惯性里程计提供了可扩展的预训练到部署范式。
原文摘要 · Abstract (English)
Robust inertial odometry is essential for various carriers when external sensing is unreliable. Learning-based methods reduce integration drift by capturing local motion priors, but these methods often remain tied to a particular carrier, limiting generalization across heterogeneous platforms. We present MosaicIMU, a carrier-conditioned Mixture-of-Experts (MoE) pretraining-and-adaptation framework for generalizable neural inertial odometry. MosaicIMU uses a prototype-based router to compose carrier-specific expert features, decodes local velocity and uncertainty constraints, and integrates them with a history-aware EKF. For unseen domain adaptation, it freezes the pretrained base model and learns a new lightweight expert residual branch. For edge-deployment, it further reuses the router to select informative online samples for efficient incremental updates. Experiments show that MosaicIMU consistently outperforms learning-based baselines, reducing average ATE and RTE-10s by 40% and 34%, respectively. These results highlight that MosaicIMU provides a scalable pretraining-to-deployment paradigm for generalizable and adaptive neural inertial odometry.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。