Description
RT-Fall is a WiFi CSI dataset for real-time fall detection by Wang et al. It captures CSI traces of fall events versus daily activities (sitting, walking, lying down) and is the canonical reference dataset for CSI-based fall detection and ambient assisted living studies.
Modality / size
- Modality: WiFi CSI from Intel 5300 NIC, 30 subcarriers.
- Subjects / scenarios: indoor settings; falls, sit-down, lie-down, walk, and other ADLs.
- Labels: binary fall vs. non-fall, often broken down per activity.
Used by (papers)
- Standard reference in CSI fall-detection literature.
- Compared against by CSI HAR methods that include falls as a critical class.
Notes
- Public dataset, released with the RT-Fall paper.