EMGBench
收藏资源简介：
EMGBench数据集由卡内基梅隆大学创建，旨在评估肌电图分类算法在分布外性能的基准。该数据集包含九个EMG数据集，涵盖了多种实验条件和数据收集协议。其中，新引入的FlexWear-HD数据集使用了一种易于穿戴的高密度EMG传感器，用于收集肌肉活动数据。数据集的创建过程包括对原始时间序列数据进行预处理，转换为2D活动图，并使用多种机器学习模型进行分类。EMGBench数据集主要应用于评估和改进EMG接口的鲁棒性和适应性，特别是在控制辅助技术如假肢和机器人方面。
The EMGBench dataset, developed by Carnegie Mellon University, serves as a benchmark for evaluating the out-of-distribution performance of electromyography (EMG) classification algorithms. It consists of nine EMG datasets spanning diverse experimental conditions and data collection protocols. Among these, the newly introduced FlexWear-HD dataset utilizes an easy-to-wear high-density EMG sensor to gather muscle activity data. The development workflow of the EMGBench dataset includes preprocessing raw time-series data, converting it into 2D activity maps, and conducting classification experiments with multiple machine learning models. The EMGBench dataset is primarily utilized to assess and enhance the robustness and adaptability of EMG interfaces, particularly for controlling assistive technologies such as prosthetics and robotic systems.
EMGBench: Benchmarking Out-of-Distribution Generalization and Adaptation for Electromyography
基本信息
- 标题: EMGBench: Benchmarking Out-of-Distribution Generalization and Adaptation for Electromyography
- 会议: NeurIPS 2024
- 作者:
- Jehan Yang
- Maxwell Soh
- Vivianna Lieu
- Douglas J. Weber
- Zackory Erickson
- 机构: Carnegie Mellon University
- 贡献声明: Jehan Yang 和 Maxwell Soh 为同等贡献
摘要
本文介绍了首个用于评估肌电图（EMG）分类算法在分布外（Out-of-Distribution, OOD）性能的机器学习基准。该基准包括两个主要任务：1) 跨受试者分类，2) 使用时间序列的训练-测试分割进行适应。该基准涵盖九个数据集，是迄今为止最大的EMG数据集集合。此外，本文还引入了一个新的数据集，该数据集使用了一种新颖的、易于穿戴的高密度EMG可穿戴设备进行数据收集。
数据集
- FlexWear-HD Dataset: 一个新引入的数据集，使用高密度EMG可穿戴设备收集数据。
链接
- arXiv: https://arxiv.org/abs/2410.23625
- 代码: https://github.com/jehanyang/emgbench
- FlexWear-HD Dataset: https://huggingface.co/datasets/jehanyang/FlexWear-HD
BibTeX
perl @misc{yang2024emgbenchbenchmarkingoutofdistributiongeneralization, title={EMGBench: Benchmarking Out-of-Distribution Generalization and Adaptation for Electromyography}, author={Jehan Yang and Maxwell Soh and Vivianna Lieu and Douglas J Weber and Zackory Erickson}, year={2024}, eprint={2410.23625}, archivePrefix={arXiv}, primaryClass={cs.LG}, url={https://arxiv.org/abs/2410.23625}, }
- 1EMGBench: Benchmarking Out-of-Distribution Generalization and Adaptation for Electromyography卡内基梅隆大学 · 2024年
