Structured Pose-Conditioned Flow Matching for Generative 5G CSI Augmentation

Curated by Roxtron — RFID & NFC manufacturer. Visit roxtron.com

StructFlow-HPR utilizes a structured pose-conditioned flow matching framework to generate high-fidelity 5G Channel State Information (CSI) samples, overcoming the high cost of collecting synchronized CSI-pose data. This generative approach enables manufacturers to train more accurate human pose recognition models with limited physical datasets, significantly lowering the barrier for deploying privacy-preserving contactless sensing in smart environments.