Classification of Driver Head Motions Using a mm-Wave FMCW Radar and Deep Convolutional Neural Network

Abstract

Eight different driver head movements are measured using a millimeter-wave FMCW radar mounted in the dashboard of a car. The micro-Doppler signatures are converted into a spectrogram image format for analysis and classification purposes. The eight different head motions exhibit unique timefrequency profiles, which can be classified by deep learning algorithms. In this study, a convolutional neural network is used to classify the eight head motions with an optimized window size. Various dataset permutations are considered, such as the effect of window width on classification accuracy and the classification accuracy of head motions in a still car compared to a moving car.

An Unobtrusive Method for Remote Quantification of Parkinson’s and Essential Tremor Using mm-Wave Sensing

“…This gives an advantage of the small size of transmitting (Tx) and receiving (Rx) antennas and hence, a compact form factor for the sensor, facilitating portability, a desired aspect for remote health monitoring. The advent of the new generation of low-cost, compact-size mmWave sensors has paved the way for various medical, military and industrial applications [48], [49], [50]. FMCW radars are continuous wave radars that transmit the signal continuously as compared to the periodic pulses in pulse radars.…”

“…According to FFT theory, the frequency resolution increases if the length of the IF signal is increased [49]. The phase of the range FFT calculated from the reflected signal changes completely by 180 degrees for every 1 mm of range change [50]. This property is crucial for estimating vibration frequencies and algorithms are required for accurate phase unwrapping [58].…”

One‐shot learning‐based driver's head movement identification using a millimetre‐wave radar sensor

“…The camera sensor is capable of recording, which is a privacy issue, and lighting conditions also have a negative impact on monitoring performance. Radars can monitor human head movements in light conditions and overcome the driver privacy issue [11,12].…”

“…Recently, artificial intelligence in general and deep learning, in particular, have achieved success in a variety of applications such as intelligent chatbots, self-driving cars, virtual assistants, speech and image recognition [13]. Taking advantage of deep learning and in order to detect the drowsy state of the driver [11,12], using the radar to collect data and apply the deep learning method to design a framework to monitor human head motion, the authors in Ref. [11] utilised a convolutional neural network (CNN) to classify four cases of the driver's head movements.…”

“…In addition, the movement of the driver's head has been analysed to infer the driver's state of drowsiness [6,7]. Human head movements were monitored using wearable sensors [6,8,9], cameras [10] and radars [11,12]. The method of using a wearable sensor has the disadvantage of being cumbersome.…”