Hand Gesture Recognition for Assisting the Elderly

With the increasing elderly population who cannot take care of themselves, the need for human-machine interaction that does not require tedious physical effort is growing. Hand gestures have the potential to provide a non-intrusive and intuitive means of interaction for Human Machine Interface (HMI)...

Full description

Saved in:
Bibliographic Details
Published in2024 3rd International Conference on Artificial Intelligence For Internet of Things (AIIoT) pp. 1 - 6
Main Authors Veena, Harin Suresh, Supriya, P
Format Conference Proceeding
LanguageEnglish
Published IEEE 03.05.2024
Subjects
Accuracy
Chatbots
CNN
Gesture recognition
Hand Gesture
NAO Robot
Radio frequency
Real-time systems
Support vector machines
SVM
Training
Online AccessGet full text
DOI10.1109/AIIoT58432.2024.10574715

Cover

More Information
Summary:With the increasing elderly population who cannot take care of themselves, the need for human-machine interaction that does not require tedious physical effort is growing. Hand gestures have the potential to provide a non-intrusive and intuitive means of interaction for Human Machine Interface (HMI). The project aims to develop a hand gesture recognition system for the NAO robot, which can respond to gestures provided by elderly people. To achieve this goal, several machine learning and deep learning models, including RF, SVM, and CNN, were compared to find the best-fitting model for hand-gesture recognition. The dataset used for training these models was obtained from Kaggle, and Python was used as the primary language for model execution. The performance of these models was evaluated using specific performance metrics with which a most optimal model is then chosen for the NAO bot. This approach has the potential to revolutionize the elderly care industry, enabling the elderly to interact with humanoid robots effortlessly, and thereby improving their quality of life.
DOI:10.1109/AIIoT58432.2024.10574715