The wavelet transform is a mathematical signal processing technique that decomposes a non-stationary signal into time-frequency representations by convolving it with scaled and translated versions of a mother wavelet function, enabling simultaneous analysis of both temporal and spectral characteristics. In CSI-based human sensing, it is particularly valuable for capturing transient, localized features in CSI amplitude and phase signals that result from human movements such as walking, gestures, or breathing, where both the timing and frequency content of motion-induced variations are diagnostically important. Key variants employed in this domain include the Discrete Wavelet Transform (DWT), which provides efficient multi-resolution decomposition suitable for feature extraction pipelines, and the Continuous Wavelet Transform (CWT), which generates scalogram representations that can be fed directly into deep learning architectures such as convolutional neural networks for end-to-end classification.
Source Papers
- A Novel Device-Free Counting Method Based on Channel Status Information ↗ — A Novel Device-Free Counting Method Based on Channel Status
- A Survey on Human Behavior Recognition Using Channel State Information ↗ — A Survey on Human Behavior Recognition Using Channel State I
- Deep Learning-Enhanced Human Sensing with Channel State Information: A Survey ↗ — Deep Learning-Enhanced Human Sensing with Channel State Info
- Device-Free Passive Identity Identification via WiFi Signals ↗ — Device-Free Passive Identity Identification via WiFi Signals
- OPERAnet, a multimodal activity recognition dataset acquired from radio frequency and vision-based sensors ↗ — OPERAnet, a multimodal activity recognition dataset acquired
- WiFi-Based Human Sensing With Deep Learning: Recent Advances, Challenges, and Opportunities ↗ — WiFi-Based Human Sensing With Deep Learning: Recent Advances