# Radar-Based Gesture Recognition Using Adaptive Top-K Selection and Multi-Stream CNNs

## TL;DR

This paper introduces a new radar-based gesture recognition system that improves accuracy by using body and hand movements together.

## Contribution

The novel Adaptive Top-K Selection preprocessing and Multi-Stream CNN architecture enhance radar gesture recognition in real-world settings.

## Key findings

- The proposed method achieved 99.5% average accuracy on the KIT FMCW gesture dataset.
- Incorporating torso and arm reflections improves recognition reliability in realistic environments.

## Abstract

With the proliferation of the Internet of Things (IoT), gesture recognition has attracted attention as a core technology in human–computer interaction (HCI). In particular, mmWave frequency-modulated continuous-wave (FMCW) radar has emerged as an alternative to vision-based approaches due to its robustness to illumination changes and advantages in privacy. However, in real-world human–machine interface (HMI) environments, hand gestures are inevitably accompanied by torso- and arm-related reflections, which can also contain gesture-relevant variations. To effectively capture these variations without discarding them, we propose a preprocessing method called Adaptive Top-K Selection, which leverages vector entropy to summarize and preserve informative signals from both hand and body reflections. In addition, we present a Multi-Stream EfficientNetV2 architecture that jointly exploits temporal range and Doppler trajectories, together with radar-specific data augmentation and a training optimization strategy. In experiments on the publicly available FMCW gesture dataset released by the Karlsruhe Institute of Technology, the proposed method achieved an average accuracy of 99.5%. These results show that the proposed approach enables accurate and reliable gesture recognition even in realistic HMI environments with co-existing body reflections.