Deep Learning Based BCI For Enhancing Quality of Life of Senior Citizens

Brain Computer Interface (BCI) reveals connection between the human brain and computers. The BCI reads the brain waves at different head locations using electrodes and converts them into actions and commands which are used further to control the computers or devices. We propose here a Deep Learning based BCI for enhancing quality of life of senior citizens. The smart, data driven system introduced here is especially useful for assisting elderly people having difficulty to interact with surrounding peripherals. The proposed system consists of two parts: an EEG processing program and microcontroller board. The digitally stored EEG recordings are used for training and testing of classifier. The accuracy of ANN model is compared with CNN model. Accuracy by CNN model is 98.3% whereas by ANN model is 66%. The classifier output can be used to switch on/off a device with the help of a calling function and a suitable controller.