Deep Learning-Based Gaze Detection System for Automobile Drivers Using a NIR Camera Sensor

Machine learning-based identification of the risk factors for postoperative nausea and vomiting in adults

Postoperative nausea and vomiting (PONV) is a common adverse effect of anesthesia. Identifying risk factors for PONV is crucial because it is associated with a longer stay in the post-anesthesia care unit, readmissions, and perioperative costs. This retrospective study used artificial intelligence to analyze data of 37,548 adult patients (aged ≥20 years) who underwent surgery under general anesthesia at Tohoku University Hospital from January 1, 2010 to December 31, 2019. To evaluate PONV, patients who experienced nausea and/or vomiting or used antiemetics within 24 hours after surgery were extracted from postoperative medical and nursing records. We create a model that predicts probability of PONV using the gradient tree boosting model, which is a widely used machine learning algorithm in many applications due to its efficiency and accuracy. The model implementation used the LightGBM framework. Data were available for 33,676 patients. Total blood loss was identified as the strongest contributor to PONV, followed by sex, total infusion volume, and patient's age. Other identified risk factors were duration of surgery (60-400 min), no blood transfusion, use of desflurane for maintenance of anesthesia, laparoscopic surgery, lateral positioning during surgery, propofol not used for maintenance of anesthesia, and epidural anesthesia at the lumbar level. The duration of anesthesia and the use of either sevoflurane or fentanyl were not identified as risk factors for PONV. We used artificial intelligence to evaluate the extent to which risk factors for PONV contribute to the development of PONV. Intraoperative total blood loss was identified as the potential risk factor most strongly associated with PONV, although it may correlate with duration of surgery, and insufficient circulating blood volume. The use of sevoflurane and fentanyl and the anesthesia time were not identified as risk factors for PONV in this study.

Comprehensive Assessment of Artificial Intelligence Tools for Driver Monitoring and Analyzing Safety Critical Events in Vehicles

Human factors are a primary cause of vehicle accidents. Driver monitoring systems, utilizing a range of sensors and techniques, offer an effective method to monitor and alert drivers to minimize driver error and reduce risky driving behaviors, thus helping to avoid Safety Critical Events (SCEs) and enhance overall driving safety. Artificial Intelligence (AI) tools, in particular, have been widely investigated to improve the efficiency and accuracy of driver monitoring or analysis of SCEs. To better understand the state-of-the-art practices and potential directions for AI tools in this domain, this work is an inaugural attempt to consolidate AI-related tools from academic and industry perspectives. We include an extensive review of AI models and sensors used in driver gaze analysis, driver state monitoring, and analyzing SCEs. Furthermore, researchers identified essential AI tools, both in academia and industry, utilized for camera-based driver monitoring and SCE analysis, in the market. Recommendations for future research directions are presented based on the identified tools and the discrepancies between academia and industry in previous studies. This effort provides a valuable resource for researchers and practitioners seeking a deeper understanding of leveraging AI tools to minimize driver errors, avoid SCEs, and increase driving safety.

Sensor-Based Classification of Primary and Secondary Car Driver Activities Using Convolutional Neural Networks

To drive safely, the driver must be aware of the surroundings, pay attention to the road traffic, and be ready to adapt to new circumstances. Most studies on driving safety focus on detecting anomalies in driver behavior and monitoring cognitive capabilities in drivers. In our study, we proposed a classifier for basic activities in driving a car, based on a similar approach that could be applied to the recognition of basic activities in daily life, that is, using electrooculographic (EOG) signals and a one-dimensional convolutional neural network (1D CNN). Our classifier achieved an accuracy of 80% for the 16 primary and secondary activities. The accuracy related to activities in driving, including crossroad, parking, roundabout, and secondary activities, was 97.9%, 96.8%, 97.4%, and 99.5%, respectively. The F1 score for secondary driving actions (0.99) was higher than for primary driving activities (0.93-0.94). Furthermore, using the same algorithm, it was possible to distinguish four activities related to activities of daily life that were secondary activities when driving a car.

Road Feature Detection for Advance Driver Assistance System Using Deep Learning

Hundreds of people are injured or killed in road accidents. These accidents are caused by several intrinsic and extrinsic factors, including the attentiveness of the driver towards the road and its associated features. These features include approaching vehicles, pedestrians, and static fixtures, such as road lanes and traffic signs. If a driver is made aware of these features in a timely manner, a huge chunk of these accidents can be avoided. This study proposes a computer vision-based solution for detecting and recognizing traffic types and signs to help drivers pave the door for self-driving cars. A real-world roadside dataset was collected under varying lighting and road conditions, and individual frames were annotated. Two deep learning models, YOLOv7 and Faster RCNN, were trained on this custom-collected dataset to detect the aforementioned road features. The models produced mean Average Precision (mAP) scores of 87.20% and 75.64%, respectively, along with class accuracies of over 98.80%; all of these were state-of-the-art. The proposed model provides an excellent benchmark to build on to help improve traffic situations and enable future technological advances, such as Advance Driver Assistance System (ADAS) and self-driving cars.

3D Gaze Estimation Using RGB-IR Cameras

In this paper, we present a framework for 3D gaze estimation intended to identify the user's focus of attention in a corneal imaging system. The framework uses a headset that consists of three cameras, a scene camera and two eye cameras: an IR camera and an RGB camera. The IR camera is used to continuously and reliably track the pupil and the RGB camera is used to acquire corneal images of the same eye. Deep learning algorithms are trained to detect the pupil in IR and RGB images and to compute a per user 3D model of the eye in real time. Once the 3D model is built, the 3D gaze direction is computed starting from the eyeball center and passing through the pupil center to the outside world. This model can also be used to transform the pupil position detected in the IR image into its corresponding position in the RGB image and to detect the gaze direction in the corneal image. This technique circumvents the problem of pupil detection in RGB images, which is especially difficult and unreliable when the scene is reflected in the corneal images. In our approach, the auto-calibration process is transparent and unobtrusive. Users do not have to be instructed to look at specific objects to calibrate the eye tracker. They need only to act and gaze normally. The framework was evaluated in a user study in realistic settings and the results are promising. It achieved a very low 3D gaze error (2.12°) and very high accuracy in acquiring corneal images (intersection over union-IoU = 0.71). The framework may be used in a variety of real-world mobile scenarios (indoors, indoors near windows and outdoors) with high accuracy.

Driver Take-Over Behaviour Study Based on Gaze Focalization and Vehicle Data in CARLA Simulator

Autonomous vehicles are the near future of the automobile industry. However, until they reach Level 5, humans and cars will share this intermediate future. Therefore, studying the transition between autonomous and manual modes is a fascinating topic. Automated vehicles may still need to occasionally hand the control to drivers due to technology limitations and legal requirements. This paper presents a study of driver behaviour in the transition between autonomous and manual modes using a CARLA simulator. To our knowledge, this is the first take-over study with transitions conducted on this simulator. For this purpose, we obtain driver gaze focalization and fuse it with the road's semantic segmentation to track to where and when the user is paying attention, besides the actuators' reaction-time measurements provided in the literature. To track gaze focalization in a non-intrusive and inexpensive way, we use a method based on a camera developed in previous works. We devised it with the OpenFace 2.0 toolkit and a NARMAX calibration method. It transforms the face parameters extracted by the toolkit into the point where the user is looking on the simulator scene. The study was carried out by different users using our simulator, which is composed of three screens, a steering wheel and pedals. We distributed this proposal in two different computer systems due to the computational cost of the simulator based on the CARLA simulator. The robot operating system (ROS) framework is in charge of the communication of both systems to provide portability and flexibility to the proposal. Results of the transition analysis are provided using state-of-the-art metrics and a novel driver situation-awareness metric for 20 users in two different scenarios.

Review and Evaluation of Eye Movement Event Detection Algorithms

Eye tracking is a technology aimed at understanding the direction of the human gaze. Event detection is a process of detecting and classifying eye movements that are divided into several types. Nowadays, event detection is almost exclusively done by applying a detection algorithm to the raw recorded eye-tracking data. However, due to the lack of a standard procedure for how to perform evaluations, evaluating and comparing various detection algorithms in eye-tracking signals is very challenging. In this paper, we used data from a high-speed eye-tracker SMI HiSpeed 1250 system and compared event detection performance. The evaluation focused on fixations, saccades and post-saccadic oscillation classification. It used sample-by-sample comparisons to compare the algorithms and inter-agreement between algorithms and human coders. The impact of varying threshold values on threshold-based algorithms was examined and the optimum threshold values were determined. This evaluation differed from previous evaluations by using the same dataset to evaluate the event detection algorithms and human coders. We evaluated and compared the different algorithms from threshold-based, machine learning-based and deep learning event detection algorithms. The evaluation results show that all methods perform well for fixation and saccade detection; however, there are substantial differences in classification results. Generally, CNN (Convolutional Neural Network) and RF (Random Forest) algorithms outperform threshold-based methods.

Single Camera Face Position-Invariant Driver's Gaze Zone Classifier Based on Frame-Sequence Recognition Using 3D Convolutional Neural Networks

Estimating the driver's gaze in a natural real-world setting can be problematic for different challenging scenario conditions. For example, faces will undergo facial occlusions, illumination, or various face positions while driving. In this effort, we aim to reduce misclassifications in driving situations when the driver has different face distances regarding the camera. Three-dimensional Convolutional Neural Networks (CNN) models can make a spatio-temporal driver's representation that extracts features encoded in multiple adjacent frames that can describe motions. This characteristic may help ease the deficiencies of a per-frame recognition system due to the lack of context information. For example, the front, navigator, right window, left window, back mirror, and speed meter are part of the known common areas to be checked by drivers. Based on this, we implement and evaluate a model that is able to detect the head direction toward these regions having various distances from the camera. In our evaluation, the 2D CNN model had a mean average recall of 74.96% across the three models, whereas the 3D CNN model had a mean average recall of 87.02%. This result show that our proposed 3D CNN-based approach outperforms a 2D CNN per-frame recognition approach in driving situations when the driver's face has different distances from the camera.

Gaze Estimation Approach Using Deep Differential Residual Network

Gaze estimation, which is a method to determine where a person is looking at given the person's full face, is a valuable clue for understanding human intention. Similarly to other domains of computer vision, deep learning (DL) methods have gained recognition in the gaze estimation domain. However, there are still gaze calibration problems in the gaze estimation domain, thus preventing existing methods from further improving the performances. An effective solution is to directly predict the difference information of two human eyes, such as the differential network (Diff-Nn). However, this solution results in a loss of accuracy when using only one inference image. We propose a differential residual model (DRNet) combined with a new loss function to make use of the difference information of two eye images. We treat the difference information as auxiliary information. We assess the proposed model (DRNet) mainly using two public datasets (1) MpiiGaze and (2) Eyediap. Considering only the eye features, DRNet outperforms the state-of-the-art gaze estimation methods with angular-error of 4.57 and 6.14 using MpiiGaze and Eyediap datasets, respectively. Furthermore, the experimental results also demonstrate that DRNet is extremely robust to noise images.

Convolutional Neural Network-Based Technique for Gaze Estimation on Mobile Devices

Eye tracking is becoming a very popular, useful, and important technology. Many eye tracking technologies are currently expensive and only available to large corporations. Some of them necessitate explicit personal calibration, which makes them unsuitable for use in real-world or uncontrolled environments. Explicit personal calibration can also be cumbersome and degrades the user experience. To address these issues, this study proposes a Convolutional Neural Network (CNN) based calibration-free technique for improved gaze estimation in unconstrained environments. The proposed technique consists of two components, namely a face component and a 39-point facial landmark component. The face component is used to extract the gaze estimation features from the eyes, while the 39-point facial landmark component is used to encode the shape and location of the eyes (within the face) into the network. Adding this information can make the network learn free-head and eye movements. Another CNN model was designed in this study primarily for the sake of comparison. The CNN model accepts only the face images as input. Different experiments were performed, and the experimental result reveals that the proposed technique outperforms the second model. Fine-tuning was also performed using the VGG16 pre-trained model. Experimental results show that the fine-tuned results of the proposed technique perform better than the fine-tuned results of the second model. Overall, the results show that 39-point facial landmarks can be used to improve the performance of CNN-based gaze estimation models.

The Identification of Non-Driving Activities with Associated Implication on the Take-Over Process

In conditionally automated driving, the engagement of non-driving activities (NDAs) can be regarded as the main factor that affects the driver's take-over performance, the investigation of which is of great importance to the design of an intelligent human-machine interface for a safe and smooth control transition. This paper introduces a 3D convolutional neural network-based system to recognize six types of driver behaviour (four types of NDAs and two types of driving activities) through two video feeds based on head and hand movement. Based on the interaction of driver and object, the selected NDAs are divided into active mode and passive mode. The proposed recognition system achieves 85.87% accuracy for the classification of six activities. The impact of NDAs on the perspective of the driver's situation awareness and take-over quality in terms of both activity type and interaction mode is further investigated. The results show that at a similar level of achieved maximum lateral error, the engagement of NDAs demands more time for drivers to accomplish the control transition, especially for the active mode NDAs engagement, which is more mentally demanding and reduces drivers' sensitiveness to the driving situation change. Moreover, the haptic feedback torque from the steering wheel could help to reduce the time of the transition process, which can be regarded as a productive assistance system for the take-over process.

Deep learning-based facial image analysis in medical research: a systematic review protocol

INTRODUCTION

Deep learning techniques are gaining momentum in medical research. Evidence shows that deep learning has advantages over humans in image identification and classification, such as facial image analysis in detecting people's medical conditions. While positive findings are available, little is known about the state-of-the-art of deep learning-based facial image analysis in the medical context. For the consideration of patients' welfare and the development of the practice, a timely understanding of the challenges and opportunities faced by research on deep-learning-based facial image analysis is needed. To address this gap, we aim to conduct a systematic review to identify the characteristics and effects of deep learning-based facial image analysis in medical research. Insights gained from this systematic review will provide a much-needed understanding of the characteristics, challenges, as well as opportunities in deep learning-based facial image analysis applied in the contexts of disease detection, diagnosis and prognosis.

METHODS

Databases including PubMed, PsycINFO, CINAHL, IEEEXplore and Scopus will be searched for relevant studies published in English in September, 2021. Titles, abstracts and full-text articles will be screened to identify eligible articles. A manual search of the reference lists of the included articles will also be conducted. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses framework was adopted to guide the systematic review process. Two reviewers will independently examine the citations and select studies for inclusion. Discrepancies will be resolved by group discussions till a consensus is reached. Data will be extracted based on the research objective and selection criteria adopted in this study.

ETHICS AND DISSEMINATION

As the study is a protocol for a systematic review, ethical approval is not required. The study findings will be disseminated via peer-reviewed publications and conference presentations.

PROSPERO REGISTRATION NUMBER

CRD42020196473.

Gaze Focalization System for Driving Applications Using OpenFace 2.0 Toolkit with NARMAX Algorithm in Accidental Scenarios

Monitoring driver attention using the gaze estimation is a typical approach used on road scenes. This indicator is of great importance for safe driving, specially on Level 3 and Level 4 automation systems, where the take over request control strategy could be based on the driver’s gaze estimation. Nowadays, gaze estimation techniques used in the state-of-the-art are intrusive and costly, and these two aspects are limiting the usage of these techniques on real vehicles. To test this kind of application, there are some databases focused on critical situations in simulation, but they do not show real accidents because of the complexity and the danger to record them. Within this context, this paper presents a low-cost and non-intrusive camera-based gaze mapping system integrating the open-source state-of-the-art OpenFace 2.0 Toolkit to visualize the driver focalization on a database composed of recorded real traffic scenes through a heat map using NARMAX (Nonlinear AutoRegressive Moving Average model with eXogenous inputs) to establish the correspondence between the OpenFace 2.0 parameters and the screen region the user is looking at. This proposal is an improvement of our previous work, which was based on a linear approximation using a projection matrix. The proposal has been validated using the recent and challenging public database DADA2000, which has 2000 video sequences with annotated driving scenarios based on real accidents. We compare our proposal with our previous one and with an expensive desktop-mounted eye-tracker, obtaining on par results. We proved that this method can be used to record driver attention databases.

RSNet: Rail semantic segmentation network for extracting aerial railroad images

In the near future, combo of UAV (Unmanned Aerial Vehicle) and computer vision will play a vital role in monitoring the condition of the railroad periodically to ensure passenger safety. The most significant module involved in railroad visual processing is obstacle detection, in which caution is obstacle fallen near track gage inside or outside. This leads to the importance of detecting and segment the railroad as three key regions, such as gage inside, rails, and background. Traditional railroad segmentation methods depend on either manual feature selection or expensive dedicated devices such as Lidar, which is typically less reliable in railroad semantic segmentation. Also, cameras mounted on moving vehicles like a drone can produce high-resolution images, so segmenting precise pixel information from those aerial images has been challenging due to the railroad surroundings chaos. RSNet is a multi-level feature fusion algorithm for segmenting railroad aerial images captured by UAV and proposes an attention-based efficient convolutional encoder for feature extraction, which is robust and computationally efficient and modified residual decoder for segmentation which considers only essential features and produces less overhead with higher performance even in real-time railroad drone imagery. The network is trained and tested on a railroad scenic view segmentation dataset (RSSD), which we have built from real-time UAV images and achieves 0.973 dice coefficient and 0.94 jaccard on test data that exhibits better results compared to the existing approaches like a residual unit and residual squeeze net.

Survey and Synthesis of State of the Art in Driver Monitoring

Road vehicle accidents are mostly due to human errors, and many such accidents could be avoided by continuously monitoring the driver. Driver monitoring (DM) is a topic of growing interest in the automotive industry, and it will remain relevant for all vehicles that are not fully autonomous, and thus for decades for the average vehicle owner. The present paper focuses on the first step of DM, which consists of characterizing the state of the driver. Since DM will be increasingly linked to driving automation (DA), this paper presents a clear view of the role of DM at each of the six SAE levels of DA. This paper surveys the state of the art of DM, and then synthesizes it, providing a unique, structured, polychotomous view of the many characterization techniques of DM. Informed by the survey, the paper characterizes the driver state along the five main dimensions—called here “(sub)states”—of drowsiness, mental workload, distraction, emotions, and under the influence. The polychotomous view of DM is presented through a pair of interlocked tables that relate these states to their indicators (e.g., the eye-blink rate) and the sensors that can access each of these indicators (e.g., a camera). The tables factor in not only the effects linked directly to the driver, but also those linked to the (driven) vehicle and the (driving) environment. They show, at a glance, to concerned researchers, equipment providers, and vehicle manufacturers (1) most of the options they have to implement various forms of advanced DM systems, and (2) fruitful areas for further research and innovation.

Does agreement mean accuracy? Evaluating glance annotation in naturalistic driving data

Naturalistic driving studies often make use of cameras to monitor driver behavior. To analyze the resulting video images, human annotation is often adopted. These annotations then serve as the 'gold standard' to train and evaluate automated computer vision algorithms, even though it is uncertain how accurate human annotation is. In this study, we provide a first evaluation of glance direction annotation by comparing instructed, actual glance direction of truck drivers with annotated direction. Findings indicate that while for some locations high annotation accuracy is achieved, for most locations accuracy is well below 50%. Higher accuracy can be obtained by clustering these locations, but this also leads to reduced detail of the annotation, suggesting that decisions to use clustering should take into account the purpose of the annotation. The data also show that high agreement between annotators does not guarantee high accuracy. We argue that the accuracy of annotation needs to be verified experimentally more often.

Driver Fatigue Detection Systems Using Multi-Sensors, Smartphone, and Cloud-Based Computing Platforms: A Comparative Analysis

Internet of things (IoT) cloud-based applications deliver advanced solutions for smart cities to decrease traffic accidents caused by driver fatigue while driving on the road. Environmental conditions or driver behavior can ultimately lead to serious roadside accidents. In recent years, the authors have developed many low-cost, computerized, driver fatigue detection systems (DFDs) to help drivers, by using multi-sensors, and mobile and cloud-based computing architecture. To promote safe driving, these are the most current emerging platforms that were introduced in the past. In this paper, we reviewed state-of-the-art approaches for predicting unsafe driving styles using three common IoT-based architectures. The novelty of this article is to show major differences among multi-sensors, smartphone-based, and cloud-based architectures in multimodal feature processing. We discussed all of the problems that machine learning techniques faced in recent years, particularly the deep learning (DL) model, to predict driver hypovigilance, especially in terms of these three IoT-based architectures. Moreover, we performed state-of-the-art comparisons by using driving simulators to incorporate multimodal features of the driver. We also mention online data sources in this article to test and train network architecture in the field of DFDs on public available multimodal datasets. These comparisons assist other authors to continue future research in this domain. To evaluate the performance, we mention the major problems in these three architectures to help researchers use the best IoT-based architecture for detecting DFDs in a real-time environment. Moreover, the important factors of Multi-Access Edge Computing (MEC) and 5th generation (5G) networks are analyzed in the context of deep learning architecture to improve the response time of DFD systems. Lastly, it is concluded that there is a research gap when it comes to implementing the DFD systems on MEC and 5G technologies by using multimodal features and DL architecture.

Pupil Localisation and Eye Centre Estimation Using Machine Learning and Computer Vision

Various methods have been used to estimate the pupil location within an image or a real-time video frame in many fields. However, these methods lack the performance specifically in low-resolution images and varying background conditions. We propose a coarse-to-fine pupil localisation method using a composite of machine learning and image processing algorithms. First, a pre-trained model is employed for the facial landmark identification to extract the desired eye frames within the input image. Then, we use multi-stage convolution to find the optimal horizontal and vertical coordinates of the pupil within the identified eye frames. For this purpose, we define an adaptive kernel to deal with the varying resolution and size of input images. Furthermore, a dynamic threshold is calculated recursively for reliable identification of the best-matched candidate. We evaluated our method using various statistical and standard metrics along with a standardised distance metric that we introduce for the first time in this study. The proposed method outperforms previous works in terms of accuracy and reliability when benchmarked on multiple standard datasets. The work has diverse artificial intelligence and industrial applications including human computer interfaces, emotion recognition, psychological profiling, healthcare, and automated deception detection.

Unobtrusive Health Monitoring in Private Spaces: The Smart Vehicle

Unobtrusive in-vehicle health monitoring has the potential to use the driving time to perform regular medical check-ups. This work intends to provide a guide to currently proposed sensor systems for in-vehicle monitoring and to answer, in particular, the questions: (1) Which sensors are suitable for in-vehicle data collection? (2) Where should the sensors be placed? (3) Which biosignals or vital signs can be monitored in the vehicle? (4) Which purposes can be supported with the health data? We reviewed retrospective literature systematically and summarized the up-to-date research on leveraging sensor technology for unobtrusive in-vehicle health monitoring. PubMed, IEEE Xplore, and Scopus delivered 959 articles. We firstly screened titles and abstracts for relevance. Thereafter, we assessed the entire articles. Finally, 46 papers were included and analyzed. A guide is provided to the currently proposed sensor systems. Through this guide, potential sensor information can be derived from the biomedical data needed for respective purposes. The suggested locations for the corresponding sensors are also linked. Fifteen types of sensors were found. Driver-centered locations, such as steering wheel, car seat, and windscreen, are frequently used for mounting unobtrusive sensors, through which some typical biosignals like heart rate and respiration rate are measured. To date, most research focuses on sensor technology development, and most application-driven research aims at driving safety. Health-oriented research on the medical use of sensor-derived physiological parameters is still of interest.

Deep Learning-Based Drivers Emotion Classification System in Time Series Data for Remote Applications

Aggressive driving emotions is indeed one of the major causes for traffic accidents throughout the world. Real-time classification in time series data of abnormal and normal driving is a keystone to avoiding road accidents. Existing work on driving behaviors in time series data have some limitations and discomforts for the users that need to be addressed. We proposed a multimodal based method to remotely detect driver aggressiveness in order to deal these issues. The proposed method is based on change in gaze and facial emotions of drivers while driving using near-infrared (NIR) camera sensors and an illuminator installed in vehicle. Driver’s aggressive and normal time series data are collected while playing car racing and truck driving computer games, respectively, while using driving game simulator. Dlib program is used to obtain driver’s image data to extract face, left and right eye images for finding change in gaze based on convolutional neural network (CNN). Similarly, facial emotions that are based on CNN are also obtained through lips, left and right eye images extracted from Dlib program. Finally, the score level fusion is applied to scores that were obtained from change in gaze and facial emotions to classify aggressive and normal driving. The proposed method accuracy is measured through experiments while using a self-constructed large-scale testing database that shows the classification accuracy of the driver’s change in gaze and facial emotions for aggressive and normal driving is high, and the performance is superior to that of previous methods.

Gaze and Eye Tracking: Techniques and Applications in ADAS

Tracking drivers’ eyes and gazes is a topic of great interest in the research of advanced driving assistance systems (ADAS). It is especially a matter of serious discussion among the road safety researchers’ community, as visual distraction is considered among the major causes of road accidents. In this paper, techniques for eye and gaze tracking are first comprehensively reviewed while discussing their major categories. The advantages and limitations of each category are explained with respect to their requirements and practical uses. In another section of the paper, the applications of eyes and gaze tracking systems in ADAS are discussed. The process of acquisition of driver’s eyes and gaze data and the algorithms used to process this data are explained. It is explained how the data related to a driver’s eyes and gaze can be used in ADAS to reduce the losses associated with road accidents occurring due to visual distraction of the driver. A discussion on the required features of current and future eye and gaze trackers is also presented.

Use of information modelling techniques to understand research trends in eye gaze estimation methods: An automated review

Eye gaze tracking has been used to study the influence of visual stimuli on consumer behavior and attentional processes. Eye gaze tracking techniques have made substantial contributions in advertisement design, human computer interaction, virtual reality and disease diagnosis. Eye gaze estimation is considered critical for prediction of human attention, and hence indispensable for better understanding human activities. In this paper, Latent Semantic Analysis is used to develop an information model for identifying emerging research trends within eye gaze estimation techniques. An exhaustive collection of 423 titles and abstracts of research papers published during 2005-2018 were used. Five major research areas and ten research trends were classified based upon this study.

Visual Saliency Prediction Based on Deep Learning

Human eye movement is one of the most important functions for understanding our surroundings. When a human eye processes a scene, it quickly focuses on dominant parts of the scene, commonly known as a visual saliency detection or visual attention prediction. Recently, neural networks have been used to predict visual saliency. This paper proposes a deep learning encoder-decoder architecture, based on a transfer learning technique, to predict visual saliency. In the proposed model, visual features are extracted through convolutional layers from raw images to predict visual saliency. In addition, the proposed model uses the VGG-16 network for semantic segmentation, which uses a pixel classification layer to predict the categorical label for every pixel in an input image. The proposed model is applied to several datasets, including TORONTO, MIT300, MIT1003, and DUT-OMRON, to illustrate its efficiency. The results of the proposed model are quantitatively and qualitatively compared to classic and state-of-the-art deep learning models. Using the proposed deep learning model, a global accuracy of up to 96.22% is achieved for the prediction of visual saliency.

Lightweight Driver Monitoring System Based on Multi-Task Mobilenets

Research on driver status recognition has been actively conducted to reduce fatal crashes caused by the driver’s distraction and drowsiness. As in many other research areas, deep-learning-based algorithms are showing excellent performance for driver status recognition. However, despite decades of research in the driver status recognition area, the visual image-based driver monitoring system has not been widely used in the automobile industry. This is because the system requires high-performance processors, as well as has a hierarchical structure in which each procedure is affected by an inaccuracy from the previous procedure. To avoid using a hierarchical structure, we propose a method using Mobilenets without the functions of face detection and tracking and show this method is enabled to recognize facial behaviors that indicate the driver’s distraction. However, frames per second processed by Mobilenets with a Raspberry pi, one of the single-board computers, is not enough to recognize the driver status. To alleviate this problem, we propose a lightweight driver monitoring system using a resource sharing device in a vehicle (e.g., a driver’s mobile phone). The proposed system is based on Multi-Task Mobilenets (MT-Mobilenets), which consists of the Mobilenets’ base and multi-task classifier. The three Softmax regressions of the multi-task classifier help one Mobilenets base recognize facial behaviors related to the driver status, such as distraction, fatigue, and drowsiness. The proposed system based on MT-Mobilenets improved the accuracy of the driver status recognition with Raspberry Pi by using one additional device.

DeepVOG: Open-source pupil segmentation and gaze estimation in neuroscience using deep learning

BACKGROUND

A prerequisite for many eye tracking and video-oculography (VOG) methods is an accurate localization of the pupil. Several existing techniques face challenges in images with artifacts and under naturalistic low-light conditions, e.g. with highly dilated pupils.

NEW METHOD

For the first time, we propose to use a fully convolutional neural network (FCNN) for segmentation of the whole pupil area, trained on 3946 VOG images hand-annotated at our institute. We integrate the FCNN into DeepVOG, along with an established method for gaze estimation from elliptical pupil contours, which we improve upon by considering our FCNN's segmentation confidence measure.

RESULTS

The FCNN output simultaneously enables us to perform pupil center localization, elliptical contour estimation and blink detection, all with a single network and with an assigned confidence value, at framerates above 130 Hz on commercial workstations with GPU acceleration. Pupil centre coordinates can be estimated with a median accuracy of around 1.0 pixel, and gaze estimation is accurate to within 0.5 degrees. The FCNN is able to robustly segment the pupil in a wide array of datasets that were not used for training.

COMPARISON WITH EXISTING METHODS

We validate our method against gold standard eye images that were artificially rendered, as well as hand-annotated VOG data from a gold-standard clinical system (EyeSeeCam) at our institute.

CONCLUSIONS

Our proposed FCNN-based pupil segmentation framework is accurate, robust and generalizes well to new VOG datasets. We provide our code and pre-trained FCNN model open-source and for free under www.github.com/pydsgz/DeepVOG.

Using Eye Tracking to Assess Gaze Concentration in Meditation

An important component of Heart Chan Meditation is gaze concentration training. Here, we determine whether eye tracking can be used to assess gaze concentration ability. Study participants (n = 306) were requested to focus their gaze on the innermost of three concentric circles for 1 min while their eye movements were recorded. Results suggest that participants with high scores on gaze concentration accuracy and precision had lower systolic blood pressure and higher sleep quality, suggesting that eye tracking may be effective to assess and train gaze concentration within Heart Chan Meditation.

Estimation of Driver's Danger Level when Accessing the Center Console for Safe Driving

This paper proposes a system for estimating the level of danger when a driver accesses the center console of a vehicle while driving. The proposed system uses a driver monitoring platform to measure the distance between the driver’s hand and the center console during driving, as well as the time taken for the driver to access the center console. Three infrared sensors on the center console are used to detect the movement of the driver’s hand. These sensors are installed in three locations: the air conditioner or heater (temperature control) button, wind direction control button, and wind intensity control button. A driver’s danger level is estimated to be based on a linear regression analysis of the distance and time of movement between the driver’s hand and the center console, as measured in the proposed scenarios. In the experimental results of the proposed scenarios, the root mean square error of driver H using distance and time of movement between the driver’s hand and the center console is 0.0043, which indicates the best estimation of a driver’s danger level.

Etracker: A Mobile Gaze-Tracking System with Near-Eye Display Based on a Combined Gaze-Tracking Algorithm

Eye tracking technology has become increasingly important for psychological analysis, medical diagnosis, driver assistance systems, and many other applications. Various gaze-tracking models have been established by previous researchers. However, there is currently no near-eye display system with accurate gaze-tracking performance and a convenient user experience. In this paper, we constructed a complete prototype of the mobile gaze-tracking system 'Etracker' with a near-eye viewing device for human gaze tracking. We proposed a combined gaze-tracking algorithm. In this algorithm, the convolutional neural network is used to remove blinking images and predict coarse gaze position, and then a geometric model is defined for accurate human gaze tracking. Moreover, we proposed using the mean value of gazes to resolve pupil center changes caused by nystagmus in calibration algorithms, so that an individual user only needs to calibrate it the first time, which makes our system more convenient. The experiments on gaze data from 26 participants show that the eye center detection accuracy is 98% and Etracker can provide an average gaze accuracy of 0.53° at a rate of 30⁻60 Hz.