Survey on Driver Drowsiness Detection System

We proposed to use this system to minimise the frequency of accidents caused by driver exhaustion, hence improving road safety. This device uses optical information and artificial intelligence to identify driver sleepiness automatically. We use Softmax to find, monitor, and analyse the driver's face and eyes in order to calculate PERCLOS (% of eye closure). It will also employ alcohol pulse detection to determine whether or not the person is normal. Due to extended driving durations and boredom in crowded settings, driver weariness is one of the leading causes of traffic accidents, particularly for drivers of big vehicles (such as buses and heavy trucks). Keywords: Driver Drowsiness, OpenCV, TensorFlow, Image Processing, Computer Vision

Adaptive Threshold Scheme for Pulse Detection under Condition of Background Nonstationary Noise

The paper describes a new adaptive threshold scheme for detecting pulses in high-frequency signals against a background of non-stationary noise. The result of the scheme operation is to determine the pulse boundaries by comparing the signal amplitude-time parameters with the threshold. The threshold value is calculated in non-overlapping windows of fixed length and depends only on the background noise level. The detected pulses undergo additional shape checking, taking into account their characteristics. The parameters of the algorithms for detecting pulses and checking their shape can be adjusted for any type of high-frequency pulse signals. This threshold scheme is tuned to detect pulses in high frequency geoacoustic emission signals. The results of the scheme operation on an artificial signal and on fragments of a geoacoustic signal are given, a comparison is made between the proposed scheme and the previously used (outdated) one. The new threshold scheme proposed by the authors is less sensitive to the choice of the initial threshold value and it is more stable in operation. When processing 15-minute fragments of a geoacoustic signal, the new scheme correctly detects, on average, 5 times more pulses.

Evaluation of an algorithm for optical pulse detection in children for application to the Pepper robot

To engage in socio-emotional interactions, children with autism spectrum conditions (ASC) need support to understand and convey emotions. In our approach, a humanoid robot (Pepper, Softbanks Robotics) acts as a tutor for the child within autism care. The robot, equipped with multimodal sensor technology to acquire the emotional feedback of the child, stimulates the child to perform tasks, adapted to its current arousal state. By in-, or decreasing the difficulties of implemented training modules, the child can be given the appropriate task according to its emotional state. The child’s arousal is measured with different techniques implemented in and on the robot: emotion detection based on audio recordings of the speech signal and camera detected facial expressions, or heart rate. To this end, the remote Photoplethysmography (rPPG) signal from camera recordings of the subjects’ face is acquired. While its unintrusive measurement is an advantage, a major drawback for rPPG is its proneness to motion and light artefacts requiring de-noising steps. A wavelet transform based on log-Gabor wavelets and a filter bank with 32 filters was implemented. The signal was filtered with a prior filter and afterwards with a Markov chain in order to extract the underlying pulse rate. Within an initial study, five children were observed watching videos with different co-notated emotions. As reference for the heart rate (HR), a wristband (empatica E4) was used. The captured emotions of all subjects were annotated to identify low and high arousal parts and positive and negative emotions. Extracted HR from rPPG-data indicated a correlation with the annotated emotions.

Arrhythmias Other Than Atrial Fibrillation in Those With an Irregular Pulse Detected With a Smartwatch: Findings From the Apple Heart Study

Background: The Apple watch irregular pulse detection algorithm was found to have a positive predictive value of 0.84 for identification of atrial fibrillation (AF). We sought to describe the prevalence of arrhythmias other than AF in those with an irregular pulse detected on a smartwatch. Methods: The Apple Heart Study investigated a smartwatch-based irregular pulse notification algorithm to identify AF. For this secondary analysis, we analyzed participants who received an ambulatory ECG patch after index irregular pulse notification. We excluded participants with AF identified on ECG patch and described the prevalence of other arrhythmias on the remaining participant ECG patches. We also reported the proportion of participants self-reporting subsequent AF diagnosis. Results: Among 419 297 participants enrolled in the Apple Heart Study, 450 participant ECG patches were analyzed, with no AF on 297 ECG patches (66%). Non-AF arrhythmias (excluding supraventricular tachycardias <30 beats and pauses <3 seconds) were detected in 119 participants (40.1%) with ECG patches without AF. The most common arrhythmias were frequent PACs (burden ≥1% to <5%, 15.8%; ≥5% to <15%, 8.8%), atrial tachycardia (≥30 beats, 5.4%), frequent PVCs (burden ≥1% to <5%, 6.1%; ≥5% to <15%, 2.7%), and nonsustained ventricular tachycardia (4–7 beats, 6.4%; ≥8 beats, 3.7%). Of 249 participants with no AF detected on ECG patch and patient-reported data available, 76 participants (30.5%) reported subsequent AF diagnosis. Conclusions: In participants with an irregular pulse notification on the Apple Watch and no AF observed on ECG patch, atrial and ventricular arrhythmias, mostly PACs and PVCs, were detected in 40% of participants. Defining optimal care for patients with detection of incidental arrhythmias other than AF is important as AF detection is further investigated, implemented, and refined.