Angle of Arrival Estimation for Gesture Recognition from reflective body-worn tags

TL;DR

This paper tackles the challenge of fine-grained hand gesture recognition using passive body-worn RFID tags by introducing AoA tracking as a discriminative feature. It combines a MUSIC-based AoA estimator with Smart Antenna Switching for fixed tags and a Kalman-smoothing tracker for moving tags to produce reliable AoA trajectories $\theta_t$. Empirical results show that integrating AoA with RSS and phase features yields substantial improvements, with accuracy gains up to about 15 percentage points and some configurations reaching $\approx 97\%$ accuracy, demonstrating the practical impact for wearable RF gesture sensing. The work highlights AoA as a robust, hardware-light enhancement for privacy-preserving gesture recognition in real-world environments.

Abstract

We investigate hand gesture recognition by leveraging passive reflective tags worn on the body. Considering a large set of gestures, distinct patterns are difficult to be captured by learning algorithms using backscattered received signal strength (RSS) and phase signals. This is because these features often exhibit similarities across signals from different gestures. To address this limitation, we explore the estimation of Angle of Arrival (AoA) as a distinguishing feature, since AoA characteristically varies during body motion. To ensure reliable estimation in our system, which employs Smart Antenna Switching (SAS), we first validate AoA estimation using the Multiple SIgnal Classification (MUSIC) algorithm while the tags are fixed at specific angles. Building on this, we propose an AoA tracking method based on Kalman smoothing. Our analysis demonstrates that, while RSS and phase alone are insufficient for distinguishing certain gesture data, AoA tracking can effectively differentiate them. To evaluate the effectiveness of AoA tracking, we implement gesture recognition system benchmarks and show that incorporating AoA features significantly boosts their performance. Improvements of up to 15% confirm the value of AoA-based enhancement.

Figures (7)

• Figure 1: Experimental setup for body-worn RFID-based gesture recognition. Two passive RFID tags are attached to the subject, with phase and RSS data collected using an Impinj Speedway R420 reader and two circularly polarized antennas.
• Figure 2: The evaluation of the MUSIC algorithm for AoA estimation is conducted for a fixed tag in the anechoic chamber ($\theta_{_\text{c}}$) and in the industrial laboratory ($\theta_\text{L}$) when the tag is located at: (a) $\theta = -5^\circ$, (b) $\theta = 0^\circ$, and (c) $\theta = 15^\circ$.
• Figure 3: The evaluation of AoA estimation with two tags, located at $-15^\circ$ and $-10^\circ$, in the anechoic chamber and the industrial laboratory.
• Figure 4: The gestures performed by participants are (a) LD: Lateral Down (1), (b) LF: Lateral to Front (2), (c) LR: Lateral Raise (3), (d) LAC: Left Arm Circle (4), (e) RAC: Right Arm Circle (5), (f) L: Lift (6), (g) Pl: Pull (7), (h) Ps: Push (8), (i) LRo: Left Round (9), (j) RR: Right Round (10), (k) SL: Swipe Left (11), (l) SR: Swipe Right (12), (m) 2HLD: Two Hands Lateral Down (13), (n) 2HLF: Two Hands Lateral to Front (14), (o) 2HLR: Two Hands Lateral Raise (15), (p) 2HIC: Two Hands Inward Circle (16), (q) 2HOC: Two Hands Outward Circle (17), (r) 2HL: Two Hands Lift (18), (s) 2HPl: Two Hands Pull (19), (t) 2HPs: Two Hands Push (20), (u) 2HR: Two Hands Round (21).
• Figure 5: Comparison of RSS, phase, and AoA features for gesture differentiation. (a) Processed RSS signals from the hand-mounted tag for a simple set of gestures. (b) Processed phase signals from the hand-mounted tag for a simple set of gestures. (c) Estimated AoA signals from the hand-mounted tag for a simple set of gestures. (d) Estimated AoA signals from the hand-mounted tags for a complex set of gestures. T1 and T2 denote Tag 1 and Tag 2, respectively.