Real-World mmWave Radar Dataset for Human Action Detection
收藏资源简介:
该数据集由Chirp公司和滑铁卢大学系统设计工程视觉与图像处理组创建,专注于使用毫米波雷达传感器在真实家庭环境中进行人体动作检测。数据集包含来自28个家庭的458次坐下和454次站起动作,覆盖48个房间,旨在解决家庭自动化和医疗保健中的隐私保护动作识别问题。数据集通过毫米波雷达和低分辨率热传感器采集,经过手动标注和数据增强处理,适用于开发基于毫米波雷达的人体动作识别模型,特别是用于评估和支持家庭环境中的移动性。
This dataset was created by Chirp Inc. and the Vision and Image Processing Group of the Department of Systems Design Engineering, University of Waterloo. It focuses on human motion detection using millimeter-wave radar sensors in real-world home environments. The dataset includes 458 instances of sitting down and 454 instances of standing up from 28 households across 48 rooms, aiming to address privacy-preserving action recognition challenges in home automation and healthcare. Collected via millimeter-wave radars and low-resolution thermal sensors, the dataset has undergone manual annotation and data augmentation, and is applicable for developing millimeter-wave radar-based human action recognition models, particularly for evaluating and supporting mobility in home environments.

- 1Dataset for Real-World Human Action Detection Using FMCW mmWave RadarChirp公司,滑铁卢大学系统设计工程视觉与图像处理组 · 2024年



