Compact Gesture Recognition Tool

Resource Overview

A lightweight gesture recognition tool developed as a Master's degree assignment, implementing block FFT algorithm with 70-80% recognition accuracy.

Detailed Documentation

I developed a straightforward gesture recognition tool capable of identifying various small hand gestures. This project was completed as a minor assignment during my Master's degree studies. The implementation utilizes a block-based Fast Fourier Transform (FFT) algorithm for gesture feature extraction and classification. The system processes input gestures by dividing them into sequential blocks, applying FFT to extract frequency-domain features, and comparing them against predefined gesture templates. Through careful parameter tuning and optimization, the tool achieves recognition accuracy between 70% and 80%. This solution enables users to perform gesture-based interactions more conveniently while providing an enhanced user experience through intuitive gesture control.