Reliability-Gated Fusion of Consumer Head and Foot IMUs for Lower-Body 3D Pose

📄 arXiv: 2609.35764v1 📥 PDF

作者: Zhilin Guo, Boqiao Zhang, Oszkár Urbán, Josef Bengtson, Hakan Aktas, Wenzhao Li, Siyu Hong, Kyle Fogarty, Chenliang Zhou, Ali Senguel, Cengiz Oztireli

分类: cs.CV, cs.HC

发布日期: 2026-09-28

备注: 10 pages, 4 figures, 3 tables. Code: https://github.com/ZhilinGuo/reliability-gated-imu-fusion

🔗 代码/项目: GITHUB

💡 一句话要点

提出可靠性门控融合方法以解决消费者IMU的低可靠性问题

🎯 匹配领域: 支柱三：空间感知与语义 (Perception & Semantics)

关键词: 姿态估计 惯性测量单元 传感器融合 运动捕捉 可靠性门控 深度学习 虚拟现实

📋 核心要点

  1. 现有的消费者IMU在姿态估计中存在可靠性问题，导致运动捕捉的准确性受到影响。
  2. 论文提出了一种基于可靠性门控的融合方法，使模型能够动态调整对不同传感器通道的信任程度。
  3. 实验结果显示，该方法在干净数据上准确性达到69.4mm，相较于其他基线有显著提升，且在各种故障情况下表现稳定。

📝 摘要（中文）

稀疏惯性姿态估计承诺实现无摄像头的运动捕捉，但消费者传感器的可靠性较低：固件融合的方向存在偏差，安装方式在不同会话中变化，数据流可能漂移或丢失。通过一个新的35次单一受试者基准，将耳塞头部惯性测量单元（IMU）与两个智能鞋垫足部IMU配对，我们发现可靠性问题是通道级别的。研究表明，脚部加速度是最具信息量的输入，而固件融合的脚部方向则是导致性能下降的因素。我们让模型学习如何信任每个流的每个通道，采用每个流每个通道块一个时间门控，并在合成损坏的预训练数据上训练辅助可靠性目标。该模型在干净数据上的准确性最高，且在每种模拟故障下表现优异。

🔬 方法详解

问题定义：本论文旨在解决消费者IMU在低可靠性下的姿态估计问题，现有方法在传感器融合时未能有效处理传感器的偏差和丢失。

核心思路：提出了一种可靠性门控机制，使模型能够根据每个通道的可靠性动态调整信任程度，从而提高姿态估计的准确性。

技术框架：整体架构包括数据采集、通道可靠性评估、门控机制和姿态融合四个主要模块。通过对每个传感器通道的历史数据进行分析，模型能够学习到各通道的可靠性。

关键创新：最重要的创新在于引入了通道门控机制，使得模型能够在测试时自动抑制不可靠的传感器通道，从而提升整体估计的准确性。

关键设计：模型采用了一个时间门控机制，每个流每个通道块都有独立的门控，损失函数中加入了辅助可靠性目标，确保模型在合成数据上进行有效预训练。

🖼️ 关键图片

fig_0
fig_1
fig_2

📊 实验亮点

实验结果表明，所提出的门控模型在干净数据上的准确性达到69.4mm，相比于静态基线（83.7mm）和未门控模型（86.6mm）有显著提升。此外，在各种故障情况下，模型的鲁棒性表现优异，能够有效抑制不可靠通道的影响。

🎯 应用场景

该研究的潜在应用领域包括运动捕捉、虚拟现实和增强现实等场景，能够为无摄像头的运动分析提供更高的准确性和可靠性。未来，随着消费者IMU技术的普及，该方法有望在智能穿戴设备中得到广泛应用，提升用户体验。

📄 摘要（原文）

Sparse inertial pose estimation promises camera-free motion capture from consumer devices, but consumer sensors are unreliable: firmware-fused orientations are biased, mounting varies between sessions, and streams drift or drop out. On a new 35-take single-subject benchmark pairing an earbud head inertial measurement unit (IMU) with two smart-insole foot IMUs (SAM-3D-Body pseudo-ground-truth labels), we show the reliability problem is channel-level: a channel ablation isolates foot acceleration as the most informative input (66.6 mm vs. 79.0 mm head-only) and the firmware-fused foot orientation as the liability that destroys the gain. We therefore let the model learn how much to trust each channel of each stream: one temporal gate per stream per channel block, trained with an auxiliary reliability objective on synthetically corrupted pretraining data. The channel-gated model is the most accurate of our learned fusion arms on clean data (69.4 mm vs. 83.7 static, 86.6 ungated) and under every simulated fault (bias in training; drift, dropout eval-only); its gates suppress the natively biased foot-orientation channels on clean real data without test-time supervision and flag dropout bursts at 0.92-0.999 AUROC. Two contrasts: dropping a channel known a priori to fail is flat across foot faults but collapses when an unanticipated stream fails (head dropout: 92.9 vs. 79.3 mm); and a fine-tuned HMD-Poser is more accurate on clean data (64.4 mm) and nominally under drift, with no significant paired difference under bias or dropout, but a larger worst-case degradation from clean (+16.1 vs. +3.5 mm, single seed). Learning to gate reliability instead of sensor count is the lever for deployable sparse inertial capture. Code is available at https://github.com/ZhilinGuo/reliability-gated-imu-fusion.