Drowsiness detector using MediaPipe + LSTM

Facial landmarks is identified through MediaPipe then EAR value is calculated. Then a LSTM model was trained to predict Drowsiness or Alert state.