R1A-PC improves human shape recovery from complex electromagnetic signals
R1A-PC: Physics-Guided Electromagnetic Inversion of Three-Dimensional Human Point Clouds in Complex Static Environments
Artificial Intelligence
Summary
It is tough to detect the exact shape of a person using electromagnetic signals in places with many obstacles like walls, because echoes from these objects can confuse the measurement. The authors propose a method called R1A-PC that compares signals taken with and without the person to highlight their shape. Their method uses physics-based calculations and special processing steps to create a detailed 3D point cloud of the person. Tests show it greatly improves accuracy compared to earlier methods, especially when removing background interference.
What this means in practice
- •For security system designers: Generate accurate 3D human images from electromagnetic signals despite environmental echoes for better intruder detection and monitoring.
- •For robotics engineers: Improve robot perception in cluttered indoor environments by reconstructing human shapes from electromagnetic measurements with higher accuracy.
Authors
Xudong Yuan, Ruyun Xu, Jingtai Yang, Xianzheng Sun
Abstract
Recovering three-dimensional human geometry from electromagnetic measure?ments in a complex static environment is difficult because strong multipath responses from walls, floors, and other objects obscure the weak target per?turbation. We propose R1A-PC, a physics-guided method that reconstructs a 2048-point human cloud from paired complex fields measured with and without the target. Complex background subtraction emphasizes target-induced ampli?tude and phase changes, while the background field remains available as an environmental condition. A frequency-balanced discrete Born adjoint produces a three-dimensional spatial knowledge map. At each of two bounded deformation stages, the decoder combines complex measurement features, background fea?tures, and multiscale physical features queried at the current point coordinates; the second stage queries again after the first coordinate update. We analyze the residual of paired subtraction, the weighted normal-operator structure of the raw adjoint, and the feasible set of the predicted cloud. In a held-out background generated by full-wave simulation under a fixed acquisition geometry, R1A-PC obtains a squared Chamfer distance of 0.001434 m2 and an F-score of 0.963080 at 0.05 m. Compared with TopNet, the Chamfer distance decreases by 70.28%. Removing physical guidance or background subtraction increases the Chamfer distance by 242.06% or 241.18%, respectively. Experiments across background layouts and poses support the complementary roles of paired subtraction and position-dependent adjoint features.