The Hampel filter is a robust outlier detection and replacement method that identifies anomalous samples in a time series by comparing each data point to the median of its neighboring values within a sliding window, flagging points that deviate beyond a threshold defined as a multiple of the local median absolute deviation (MAD), and replacing detected outliers with the local median. In CSI-based sensing research, it is used during signal preprocessing to suppress impulsive noise and sporadic measurement errors inherent in raw CSI amplitude and phase streams, thereby improving the quality of features fed into downstream activity recognition or occupancy estimation pipelines. The primary variant involves tuning two parameters — the window half-width and the MAD multiplier threshold — to balance sensitivity to genuine outliers against false removal of legitimate signal transients, with stricter thresholds commonly applied in low-mobility occupancy scenarios and more permissive settings used where rapid signal fluctuations are expected during dynamic human behavior recognition tasks.
Source Papers
- A Survey on Human Behavior Recognition Using Channel State Information ↗ — A Survey on Human Behavior Recognition Using Channel State I
- CSI-Based NTC Using Ambient WiFi: Channel Selection, Topology Control and Traffic Interference ↗ — CSI-Based NTC Using Ambient WiFi: Channel Selection, Topolog
- Implementing Wi-Fi CSI-based room-level occupancy Estimation: an experimental study in multi-zone residential environments ↗ — Implementing Wi-Fi CSI-based room-level occupancy Estimation
- Towards Energy Efficient Wireless Sensing by Leveraging Ambient Wi-Fi Traffic ↗ — Towards Energy Efficient Wireless Sensing by Leveraging Ambi
- Towards Environment Independent Device Free Human Activity Recognition ↗ — Towards Environment Independent Device Free Human Activity R
- Wi-CaL: WiFi Sensing and Machine Learning Based Device-Free Crowd Counting and Localization ↗ — Wi-CaL: WiFi Sensing and Machine Learning Based Device-Free
- WiFi Sensing with Channel State Information ↗ — WiFi Sensing with Channel State Information