# Wi-Fringe: Leveraging Text Semantics in WiFi CSI-Based Device-Free Named Gesture Recognition

## TL;DR

Wi-Fringe is a WiFi CSI-based system that recognizes semantically meaningful named gestures and activities, even with minimal or no training data, by leveraging text semantics.

## Contribution

It introduces a novel approach that combines text semantics with WiFi CSI data to improve gesture recognition with limited training samples.

## Key findings

- Can recognize named gestures with few or no training examples
- Achieves accurate detection of activities at runtime
- Reduces the need for extensive training data in WiFi-based recognition

## Abstract

The lack of adequate training data is one of the major hurdles in WiFi-based activity recognition systems. In this paper, we propose Wi-Fringe, which is a WiFi CSI-based device-free human gesture recognition system that recognizes named gestures, i.e., activities and gestures that have a semantically meaningful name in English language, as opposed to arbitrary free-form gestures. Given a list of activities (only their names in English text), along with zero or more training examples (WiFi CSI values) per activity, Wi-Fringe is able to detect all activities at runtime. In other words, a subset of activities that Wi-Fringe detects do not require any training examples at all.