Machine Learning Regression Models for Real-Time Touchless Interaction Applications

Touchless technologies have gained significant popularity, particularly amidst the COVID-19 pandemic, as they addressed concerns related to germ transmission and hygiene during human-device interactions. This study aimed to develop an intuitive and user-friendly touchless system by combining eye gaze and hand gesture methods. Four regression machine learning models, namely Ridge regression, Lasso regression, Linear regression, and Gradient-boosting regressor, were trained and tested using standard metrics such as coefficient of determination (R2), mean absolute error (MAE), and mean squared error (MSE). The results indicated that Ridge regression outperformed the other models, demonstrating a high R2 value of 0.98. Leveraging this model, a simulation was conducted to evaluate the effectiveness of integrating eye gaze and hand gestures in real-time touchless interactions. The simulation demonstrated the successful integration of these modalities in the item selection process, providing users with a seamless and intuitive interaction method. This touchless interaction technology enabled effortless and accurate navigation of screens and facilitated item selections. The promising results highlighted the potential of this technology, presenting exciting opportunities for its integration with actuators. By incorporating actuators, touchless interaction systems could revolutionize various industries, including retail, healthcare, hospitality, and smart home automation.