PGUDA: Pressure-Guided Unsupervised Domain Adaptation with Cross-Modal Knowledge Distillation for sEMG-Based Gesture Recognition

📄 arXiv: 2606.31349v1 📥 PDF

作者: Yurui Liu, Xiao-Cong Zhong, Qisong Wang, Xuefu Wang, Dan Liu, Jinwei Sun

分类: eess.SP, cs.AI

发布日期: 2026-06-30

💡 一句话要点

提出PGUDA框架以解决sEMG手势识别中的领域适应问题

🎯 匹配领域: 支柱二：RL算法与架构 (RL & Architecture) 支柱九：具身大模型 (Embodied Foundation Models)

关键词: sEMG手势识别 无监督领域适应 跨模态知识蒸馏 压力信号 特征对齐 深度学习 人机交互

📋 核心要点

  1. 现有sEMG手势识别方法在不同受试者和会话间存在特征分布差异，导致性能显著下降。
  2. 本文提出PGUDA框架，通过压力信号引导无监督领域适应，利用跨模态知识蒸馏提升特征对齐效果。
  3. 实验结果表明，PGUDA在分类任务中平均准确率达到58.08%，在标注数据利用效率上表现优异。

📝 摘要（中文）

基于表面肌电图（sEMG）的手势识别在自然人机交互中展现出良好前景，但由于不同受试者和记录会话间特征分布差异，实际应用面临挑战。传统领域适应（DA）方法在对齐sEMG特征时效果不佳，主要由于其固有的随机性和标注数据稀缺。为此，本文提出了一种新颖的压力引导无监督领域适应（PGUDA）框架，利用压力信号的稳健性和稳定性，提出跨模态知识蒸馏策略，传递一致的物理语义。通过在无标注目标域上，利用在压力信号上训练的教师网络指导sEMG学生网络，正则化表示学习过程。大量实验验证了PGUDA在跨受试者和跨会话分类任务中的有效性，平均准确率达到58.08%，显著优于现有DA方法，并且在仅使用5%标注数据的情况下，分类准确率与完全监督基准相当。

🔬 方法详解

问题定义：本文旨在解决sEMG手势识别中由于不同受试者和记录会话导致的特征分布差异问题。现有领域适应方法在对齐sEMG特征时效果不佳，主要由于其固有的随机性和标注数据的稀缺性。

核心思路：PGUDA框架的核心思路是利用压力信号的稳健性，通过跨模态知识蒸馏策略，将一致的物理语义从压力信号转移到sEMG特征上，从而提升特征对齐的效果。

技术框架：PGUDA框架包括两个主要模块：教师网络和学生网络。教师网络在压力信号上进行训练，学生网络则在无标注的目标域上进行学习。教师网络指导学生网络的表示学习过程，确保知识的转移和对齐。

关键创新：PGUDA的主要创新在于引入了压力信号作为教师网络，利用其稳定性来指导sEMG特征的学习。这一方法与传统DA方法的本质区别在于，前者通过跨模态知识蒸馏实现了更有效的特征对齐。

关键设计：在网络结构上，教师网络和学生网络均采用深度学习模型，损失函数设计为结合了知识蒸馏和对抗训练的复合损失，以增强特征的可迁移性和模态不变性。

🖼️ 关键图片

fig_0
fig_1
fig_2

📊 实验亮点

PGUDA在跨受试者和跨会话分类任务中表现出色，平均准确率达到58.08%，显著优于现有领域适应方法。此外，PGUDA在标注数据利用效率上表现优异，仅需5%的标注数据即可达到与完全监督基准相当的分类准确率。

🎯 应用场景

该研究的潜在应用领域包括人机交互、康复医疗和智能控制等。通过提升sEMG手势识别的准确性和效率，PGUDA框架能够在实际应用中显著降低校准负担，推动相关技术的普及与发展。

📄 摘要（原文）

Surface electromyography (sEMG)-based gesture recognition has emerged as a promising technology for natural human-computer interaction. However, its practical deployment remains challenging due to severe performance degradation caused by feature distribution discrepancies across different subjects and recording sessions. Although domain adaptation (DA) techniques are commonly employed to mitigate such discrepancies, conventional methods often struggle to effectively aligning sEMG features, primarily due to their inherent stochasticity and the scarcity of labeled data. To address these limitations, this paper proposes a novel Pressure-Guided Unsupervised Domain Adaptation (PGUDA) framework, which leverages the robustness and stability of pressure signals to introduce a cross-modal knowledge distillation strategy that transfers consistent physical semantics across modalities. Specifically, a teacher network trained on pressure signals guides an sEMG student network on unlabeled target domains, thereby regularizing the representation learning process with transferable and modality-invariant knowledge. Extensive experiments conducted on a self-collected multimodal dataset involving eleven subjects validate the effectiveness of the proposed PGUDA framework. The results demonstrate that our proposed PGUDA achieves leading performance in both cross-subject and cross-session classification tasks, achieving average accuracies of 58.08% and substantially outperforming existing DA approaches. Notably, PGUDA exhibits remarkable label efficiency: it attains classification accuracy comparable to fully supervised benchmarks while requiring only 5% of labeled data for teacher network training. This framework offers a robust and data-efficient solution that can significantly reduce the calibration burden in practical sEMG-based gesture recognition systems.