Technology Category
Intelligent Robots: State EstimationSearch and Optimization: Learning to SearchPlanning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
DWSFormer: A Lightweight Inertial Odometry Network for Complex Motion Modeling
🤖 AI Summary
To address severe drift and abrupt accuracy degradation of inertial odometry (IO) under complex motions—e.g., sharp turns—this paper proposes a lightweight neural network framework. Methodologically: (i) a Star Operation is introduced to implicitly map inputs into a high-dimensional nonlinear feature space; (ii) a cross-channel–temporal collaborative attention mechanism is designed to enhance modeling of dynamic interdependencies across multi-source inertial signals; and (iii) a multi-scale gated convolutional unit is proposed to precisely capture fine-grained temporal variations in motion dynamics. Evaluated on six mainstream inertial datasets, the method consistently outperforms state-of-the-art approaches. On the RoNIN dataset, it achieves an absolute trajectory error (ATE) of 65.78%, representing a 2.26% relative reduction over prior work. The framework significantly improves localization robustness in challenging scenarios and establishes a new performance benchmark for learning-based inertial odometry.
Technology Category
Application Category
📝 Abstract
Inertial odometry (IO) directly estimates the position of a carrier from inertial sensor measurements and serves as a core technology for the widespread deployment of consumer grade localization systems. While existing IO methods can accurately reconstruct simple and near linear motion trajectories, they often fail to account for drift errors caused by complex motion patterns such as turning. This limitation significantly degrades localization accuracy and restricts the applicability of IO systems in real world scenarios. To address these challenges, we propose a lightweight IO framework. Specifically, inertial data is projected into a high dimensional implicit nonlinear feature space using the Star Operation method, enabling the extraction of complex motion features that are typically overlooked. We further introduce a collaborative attention mechanism that jointly models global motion dynamics across both channel and temporal dimensions. In addition, we design Multi Scale Gated Convolution Units to capture fine grained dynamic variations throughout the motion process, thereby enhancing the model's ability to learn rich and expressive motion representations. Extensive experiments demonstrate that our proposed method consistently outperforms SOTA baselines across six widely used inertial datasets. Compared to baseline models on the RoNIN dataset, it achieves reductions in ATE ranging from 2.26% to 65.78%, thereby establishing a new benchmark in the field.
Problem
Research questions and friction points this paper is trying to address.
Addresses drift errors in inertial odometry during complex motions
Improves localization accuracy for non-linear motion patterns
Enhances motion feature extraction in lightweight IO systems
Innovation
Methods, ideas, or system contributions that make the work stand out.
Star Operation for high-dimensional feature projection
Collaborative attention for global motion modeling
Multi-scale gated units for dynamic variation capture
🔎 Similar Papers
No similar papers found.
