Virtual Booth

A stroke, also called cerebrovascular accident, impairs blood supply to the brain, leading to cell death and resulting in physical as well as speech impairments. Even with the current assistive devices for Augmentative and Alternative Communication (AAC), patients still require help from someone else for basic tasks. To attend to this, an integrated room system that employs machine learning to interpret hand movements was established. The lamp, fan, system readiness, calling a helper or fan speed are controlled by six specific gestures only. The project process includes data collection sessions with gesture classification being one of its stages while hardware development is another. A deep learning model using Keras LSTM has 97% accuracy on interpreting human gestures when it is trained with 500 instances per gesture. By employing Arduino UNO for output control, the system becomes a dependable tool used by victims of stroke to effectively manage room appliances according to their own wills.