TY - GEN
T1 - Driver Drowsiness Detection Using Facial Parameters and RNNs with LSTM
AU - Yarlagadda, Vishnu
AU - Koolagudi, Shashidhar G.
AU - Kumar M V, Manoj
AU - Donepudi, Swapna
N1 - Publisher Copyright:
© 2020 IEEE.
PY - 2020/12/10
Y1 - 2020/12/10
N2 - The drowsiness is an intermediate state between awake and sleep, in which the observation and analysis of a conductor is very small. The lack of concentration due to the driver fatigue is a major cause that leads to the high number of accidents. In this work, an effort has been put to detect the state of drowsiness using facial parameters obtained using facial points. Moreover, the parameters related to eye and mouth organs have also been extracted. Deep neural networks are outperforming when compared to many state-of-the art algorithms. Hence, recurrent neural networks (RNNs) and long short-term memory (LSTM) units are considered to estimate the drowsiness level of a driver. It is found that they are very appropriate in processing of sequential multimedia data. An accuracy of 97.25% is obtained with the proposed approach.
AB - The drowsiness is an intermediate state between awake and sleep, in which the observation and analysis of a conductor is very small. The lack of concentration due to the driver fatigue is a major cause that leads to the high number of accidents. In this work, an effort has been put to detect the state of drowsiness using facial parameters obtained using facial points. Moreover, the parameters related to eye and mouth organs have also been extracted. Deep neural networks are outperforming when compared to many state-of-the art algorithms. Hence, recurrent neural networks (RNNs) and long short-term memory (LSTM) units are considered to estimate the drowsiness level of a driver. It is found that they are very appropriate in processing of sequential multimedia data. An accuracy of 97.25% is obtained with the proposed approach.
UR - https://www.scopus.com/pages/publications/85101518691
UR - https://www.scopus.com/pages/publications/85101518691#tab=citedBy
U2 - 10.1109/INDICON49873.2020.9342348
DO - 10.1109/INDICON49873.2020.9342348
M3 - Conference contribution
AN - SCOPUS:85101518691
T3 - 2020 IEEE 17th India Council International Conference, INDICON 2020
BT - 2020 IEEE 17th India Council International Conference, INDICON 2020
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 17th IEEE India Council International Conference, INDICON 2020
Y2 - 10 December 2020 through 13 December 2020
ER -