Seeing Through WiFi: Lightweight Human Pose Estimation with Dynamic Kernel Attention
2026-07-03 • Computer Vision and Pattern Recognition
Computer Vision and Pattern RecognitionMachine Learning
AI summaryⓘ
The authors developed WiLHPE, a new system that uses WiFi signals to estimate human body poses without cameras or wearable devices, helping protect privacy. Their approach trains with camera data but works using raw WiFi signals to guess poses, using a special neural network that adapts its filters and focuses on important features efficiently. They also used an advanced method to quickly find the best settings for their network. Tests show WiLHPE works better than previous methods on standard datasets and stays accurate even when the WiFi signals are noisy.
WiFi CSIhuman pose estimationneural networksattention mechanismsconvolutional kernelsTree-Structured Parzen EstimatorPCK50signal noisemodel efficiencyhyperparameter optimization
Authors
Toan D. Gian, Van-Dinh Nguyen, Vo Phi Son, Nhan Thanh Nguyen, Dinh Thai Hoang, Diep N. Nguyen, Nguyen Cong Luong, Symeon Chatzinotas
Abstract
WiFi-based human pose estimation (HPE) enables the detection and interpretation of human body positions and movements without the need for wearable devices while preserving individual privacy concerns. Implementing this solution requires enhancing model performance and maintaining efficiency, especially on resource-constrained devices. This paper introduces a novel framework, WiLHPE, for lightweight and efficient human pose estimation using WiFi CSI signals. Empowered by a camera-based model during training, WiLHPE processes raw WiFi signals directly to estimate human poses in the testing phase. It employs a novel neural network architecture to dynamically learn convolutional kernels and apply attention mechanisms across channel and frequency spaces. This innovative method diversifies the kernels to improve the recognition capabilities of WiFi signals without adding complexity, ensuring efficiency. Additionally, the Tree-Structured Parzen Estimator algorithm is employed to optimize the critical hyperparameters of the neural network efficiently, minimizing the time required for optimal hyperparameter search compared to heuristic methods. Results from experiments on both the MM-Fi and WiPose datasets highlight the superiority of WiLHPE over state-of-the-art approaches, achieving 85.96% and 94.27% at PCK50, respectively, with minimal computational overhead. Notably, WiLHPE performs impressively even under challenging conditions, maintaining around 80% at PCK50 under AWGN noise with an error variance of 0.5.