scholarly journals Lane marking detection using simple encode decode deep learning technique: SegNet

2021 ◽  
Vol 11 (4) ◽  
pp. 3032
Author(s):  
A. Al Mamun ◽  
P. P. Em ◽  
J. Hossen
Keyword(s):  
Deep Learning ◽  
Computational Complexity ◽  
False Negative ◽  
Road Accidents ◽  
Financial Loss ◽  
Driver Assistance ◽  
Driver Assistance Systems ◽  
Low Light ◽  
Lane Marking ◽  
Deep Learning Model

<span>In recent times, many innocent people are suffering from sudden death for the sake of unwanted road accidents, which also riveting a lot of financial properties. The researchers have deployed advanced driver assistance systems (ADAS) in which a large number of automated features have been incorporated in the modern vehicles to overcome human mortality as well as financial loss, and lane markings detection is one of them. Many computer vision techniques and intricate image processing approaches have been used for detecting the lane markings by utilizing the handcrafted with highly specialized features. However, the systems have become more challenging due to the computational complexity, overfitting, less accuracy, and incapability to cope up with the intricate environmental conditions. Therefore, this research paper proposed a simple encode-decode deep learning model to detect lane markings under the distinct environmental condition with lower computational complexity. The model is based on SegNet architecture for improving the performance of the existing researches, which is trained by the lane marking dataset containing different complex environment conditions like rain, cloud, low light, curve roads. The model has successfully achieved 96.38% accuracy, 0.0311 false positive, 0.0201 false negative, 0.960 F1 score with a loss of only 1.45%, less overfitting and 428 ms per step that outstripped some of the existing researches. It is expected that this research will bring a significant contribution to the field lane marking detection.</span>

scholarly journals A Review of Deep Learning-Based Methods for Pedestrian Trajectory Prediction

Sensors ◽  
2021 ◽  
Vol 21 (22) ◽  
pp. 7543
Author(s):  
Bogdan Ilie Sighencea ◽  
Rareș Ion Stanciu ◽  
Cătălin Daniel Căleanu
Keyword(s):  
Deep Learning ◽  
Performance Metrics ◽  
Evaluation Process ◽  
Surveillance Systems ◽  
Driver Assistance ◽  
Driver Assistance Systems ◽  
Trajectory Prediction ◽  
Processing Technologies ◽  
Practical Applications ◽  
New Research

Pedestrian trajectory prediction is one of the main concerns of computer vision problems in the automotive industry, especially in the field of advanced driver assistance systems. The ability to anticipate the next movements of pedestrians on the street is a key task in many areas, e.g., self-driving auto vehicles, mobile robots or advanced surveillance systems, and they still represent a technological challenge. The performance of state-of-the-art pedestrian trajectory prediction methods currently benefits from the advancements in sensors and associated signal processing technologies. The current paper reviews the most recent deep learning-based solutions for the problem of pedestrian trajectory prediction along with employed sensors and afferent processing methodologies, and it performs an overview of the available datasets, performance metrics used in the evaluation process, and practical applications. Finally, the current work exposes the research gaps from the literature and outlines potential new research directions.

scholarly journals AFFECTIVE-SOUND EFFECTS ON DRIVING BEHAVIOUR

Transport ◽  
2013 ◽  
Vol 29 (1) ◽  
pp. 100-106 ◽  
Author(s):  
Jesús Serrano ◽  
Leandro Luigi Di Stasi ◽  
Alberto Megías ◽  
Andrés Catena
Keyword(s):  
Driving Safety ◽  
Auditory Signal ◽  
Road Accidents ◽  
Driver Assistance ◽  
Driver Assistance Systems ◽  
Emotional Quality ◽  
Sound Effects ◽  
Technological Developments ◽  
Assistance Systems

Recent technological developments in active advanced driver assistance systems and in-car infotainment devices have contributed to reducing the number and severity of road accidents as well as improving and simplifying driver experience. However, these systems may impact driving performance in undesired ways, especially when emotionally-charged stimuli are used as warning signals. Emotional distraction can be a serious danger, causing delays in information processing, and reducing driving safety below minimal acceptable levels. Here we study the effect of emotionally-laden auditory signals on the speed of concurrent driving decisions. We distinguished two categories of behavioural responses: ‘urgent’ vs ‘evaluative’. In the experiments reported here participants were quicker to evaluate whether a traffic scene was risky or not after hearing an emotionally-charged auditory stimulus than after a neutral one. However, urgent (braking) responses to the same scenes were not affected by the emotional quality of the auditory signal. Based on these results, we give preliminary advice on the design of guidelines for in-car interfaces particularly in the field of affective in-car computing.

scholarly journals Can ADAS Distract Driver’s Attention? An RGB-D Camera and Deep Learning-Based Analysis

2021 ◽  
Vol 11 (24) ◽  
pp. 11587
Author(s):  
Luca Ulrich ◽  
Francesca Nonis ◽  
Enrico Vezzetti ◽  
Sandro Moos ◽  
Giandomenico Caruso ◽  
...  
Keyword(s):  
Deep Learning ◽  
Expression Recognition ◽  
Driver Assistance ◽  
Driver Assistance Systems ◽  
Driving Experience ◽  
Vehicle Accidents ◽  
Non Invasive ◽  
Driver Inattention ◽  
Assistance Systems ◽  
Invasive Techniques

Driver inattention is the primary cause of vehicle accidents; hence, manufacturers have introduced systems to support the driver and improve safety; nonetheless, advanced driver assistance systems (ADAS) must be properly designed not to become a potential source of distraction for the driver due to the provided feedback. In the present study, an experiment involving auditory and haptic ADAS has been conducted involving 11 participants, whose attention has been monitored during their driving experience. An RGB-D camera has been used to acquire the drivers’ face data. Subsequently, these images have been analyzed using a deep learning-based approach, i.e., a convolutional neural network (CNN) specifically trained to perform facial expression recognition (FER). Analyses to assess possible relationships between these results and both ADAS activations and event occurrences, i.e., accidents, have been carried out. A correlation between attention and accidents emerged, whilst facial expressions and ADAS activations resulted to be not correlated, thus no evidence that the designed ADAS are a possible source of distraction has been found. In addition to the experimental results, the proposed approach has proved to be an effective tool to monitor the driver through the usage of non-invasive techniques.

Well Connectivity Analysis with Deep Learning

2021 ◽  
Author(s):  
Yuanjun Li ◽  
Satomi Suzuki ◽  
Roland Horne
Keyword(s):  
Neural Network ◽  
Time Series ◽  
Deep Learning ◽  
False Negative ◽  
Signal Propagation ◽  
Learning Model ◽  
Short Term ◽  
Production Wells ◽  
Interference Tests ◽  
Deep Learning Model

Abstract Knowledge of well connectivity in a reservoir is crucial, especially for early-stage field development and water injection management. However, traditional interference tests can often take several weeks or even longer depending on the distance between wells and the hydraulic diffusivity of the reservoir. Therefore, instead of physically shutting in production wells, we can take advantage of deep learning methods to perform virtual interference tests. In this study, we first used the historical field data to train the deep learning model, a modified Long- and Short-term Time-series network (LSTNet). This model combines the Convolution Neural Network (CNN) to extract short-term local dependency patterns, the Recurrent Neural Network (RNN) to discover long-term patterns for time series trends, and a traditional autoregressive model to alleviate the scale insensitive problem. To address the time-lag issue in signal propagation, we employed a skip-recurrent structure that extends the existing RNN structure by connecting a current state with a previous state when the flow rate signal from an adjacent well starts to impact the observation well. In addition, we found that wells connected to the same manifold usually have similar liquid production patterns, which can lead to false causation of subsurface pressure communication. Thus we enhanced the model performance by using external feature differences to remove the surface connection in the data, thereby reducing input similarity. This enhancement can also amplify the weak signal and thus distinguish input signals. To examine the deep learning model, we used the datasets generated from Norne Field with two different geological settings: sealing and nonsealing cases. The production wells are placed at two sides of the fault to test the false-negative prediction. With these improvements and with parameter tuning, the modified LSTNet model could successfully indicate the well connectivity for the nonsealing cases and reveal the sealing structures in the sealing cases based on the historical data. The deep learning method we employed in this work can predict well pressure without using hand-crafted features, which are usually formed based on flow patterns and geological settings. Thus, this method should be applicable to general cases and more intuitive. Furthermore, this virtual interference test with a deep learning framework can avoid production loss.

Imaging lidar prototype with homography and deep learning ranging methods

2022 ◽  
Author(s):  
Sehyeon Kim ◽  
Zhaowei Chen ◽  
Hossein Alisafaee
Keyword(s):  
Deep Learning ◽  
Autonomous Vehicles ◽  
Optical Element ◽  
Driver Assistance ◽  
Driver Assistance Systems ◽  
Lidar System ◽  
Time Performance ◽  
Potential Applications ◽  
Assistance Systems ◽  
Final System

Abstract We report on developing a non-scanning laser-based imaging lidar system based on a diffractive optical element with potential applications in advanced driver assistance systems, autonomous vehicles, drone navigation, and mobile devices. Our proposed lidar utilizes image processing, homography, and deep learning. Our emphasis in the design approach is on the compactness and cost of the final system for it to be deployable both as standalone and complementary to existing lidar sensors, enabling fusion sensing in the applications. This work describes the basic elements of the proposed lidar system and presents two potential ranging mechanisms, along with their experimental results demonstrating the real-time performance of our first prototype.

scholarly journals An efficient encode-decode deep learning network for lane markings instant segmentation

2021 ◽  
Vol 11 (6) ◽  
pp. 4982
Author(s):  
A. Al Mamun ◽  
P. P. Em ◽  
J. Hossen
Keyword(s):  
Deep Learning ◽  
Environmental Conditions ◽  
False Negative ◽  
Distinct Type ◽  
Weather Conditions ◽  
Driver Assistance Systems ◽  
Traffic Condition ◽  
Computational Framework ◽  
Proposed Model ◽  
Deep Learning Network

<span lang="EN-US">Nowadays, advanced driver assistance systems (ADAS) has been incorporated with a distinct type of progressive and essential features. One of the most preliminary and significant features of the ADAS is lane marking detection, which permits the vehicle to keep in a particular road lane itself. It has been detected by utilizing high-specialized, handcrafted features and distinct post-processing approaches lead to less accurate, less efficient, and high computational framework under different environmental conditions. Hence, this research proposed a simple encode-decode deep learning approach under distinguishing environmental effects like different daytime, multiple lanes, different traffic condition, good and medium weather conditions for detecting the lane markings more accurately and efficiently. The proposed model is emphasized on the simple encode-decode Seg-Net framework incorporated with VGG16 architecture that has been trained by using the inequity and cross-entropy losses to obtain more accurate instant segmentation result of lane markings. The framework has been trained and tested on a vast public dataset named Tusimple, which includes around 3.6K training and 2.7 k testing image frames of different environmental conditions. The model has noted the highest accuracy, 96.61%, F1 score 96.34%, precision 98.91%, and recall 93.89%. Also, it has <span>also obtained the lowest 3.125% false positive and 1.259% false-negative value, which transcended some of the previous researches. It is expected to</span> assist significantly in the field of lane markings detection applying deep neural networks.</span>

scholarly journals A Hybrid Deep Learning Model for Recognizing Actions of Distracted Drivers

Sensors ◽  
2021 ◽  
Vol 21 (21) ◽  
pp. 7424
Author(s):  
Shuang-Jian Jiao ◽  
Lin-Yao Liu ◽  
Qian Liu
Keyword(s):  
Deep Learning ◽  
Human Body ◽  
Traffic Accidents ◽  
Clustering Algorithm ◽  
Short Term Memory ◽  
Learning Model ◽  
Cluster Center ◽  
Navigation Systems ◽  
Driver Assistance Systems ◽  
Deep Learning Model

With the rapid spreading of in-vehicle information systems such as smartphones, navigation systems, and radios, the number of traffic accidents caused by driver distractions shows an increasing trend. Timely identification and warning are deemed to be crucial for distracted driving and the establishment of driver assistance systems is of great value. However, almost all research on the recognition of the driver’s distracted actions using computer vision methods neglected the importance of temporal information for action recognition. This paper proposes a hybrid deep learning model for recognizing the actions of distracted drivers. Specifically, we used OpenPose to obtain skeleton information of the human body and then constructed the vector angle and modulus ratio of the human body structure as features to describe the driver’s actions, thereby realizing the fusion of deep network features and artificial features, which improve the information density of spatial features. The K-means clustering algorithm was used to preselect the original frames, and the method of inter-frame comparison was used to obtain the final keyframe sequence by comparing the Euclidean distance between manually constructed vectors representing frames and the vector representing the cluster center. Finally, we constructed a two-layer long short-term memory neural network to obtain more effective spatiotemporal features, and one softmax layer to identify the distracted driver’s action. The experimental results based on the collected dataset prove the effectiveness of this framework, and it can provide a theoretical basis for the establishment of vehicle distraction warning systems.

Sign in / Sign up

Export Citation Format

Share Document