Few-shot Domain Adaptation for IMU Denoising

TL;DR

This paper introduces a few-shot domain adaptation framework for IMU denoising, effectively handling different error characteristics across scenarios with limited data, demonstrated on multiple datasets and robots.

Contribution

The work presents a novel domain adaptation framework with a reconstitution loss and few-shot training strategy for IMU denoising across diverse scenarios.

Findings

Effective denoising performance shown in orientation results

Framework verified by t-SNE visualization of domain adaptability

Demonstrated on multiple datasets and robot platforms

Abstract

Different application scenarios will cause IMU to exhibit different error characteristics which will cause trouble to robot application. However, most data processing methods need to be designed for specific scenario. To solve this problem, we propose a few-shot domain adaptation method. In this work, a domain adaptation framework is considered for denoising the IMU, a reconstitution loss is designed to improve domain adaptability. In addition, in order to further improve the adaptability in the case of limited data, a few-shot training strategy is adopted. In the experiment, we quantify our method on two datasets (EuRoC and TUM-VI) and two real robots (car and quadruped robot) with three different precision IMUs. According to the experimental results, the adaptability of our framework is verified by t-SNE. In orientation results, our proposed method shows the great denoising performance.