Academic literature on the topic 'Deep learning, computer vision, safety, road scene understanding'

In the past decade, autonomous vehicle systems (AVS) have advanced at an exponential rate, particularly due to improvements in artificial intelligence, which have had a significant impact on social as well as road safety and the future of transportation systems. However, the AVS is still far away from mass production because of the high cost of sensor fusion and a lack of combination of top-tier solutions to tackle uncertainty on roads. To reduce sensor dependency and to increase manufacturing along with enhancing research, deep learning-based approaches could be the best alternative for developing practical AVS. With this vision, in this systematic review paper, we broadly discussed the literature of deep learning for AVS from the past decade for real-life implementation in core fields. The systematic review on AVS implementing deep learning is categorized into several modules that cover activities including perception analysis (vehicle detection, traffic signs and light identification, pedestrian detection, lane and curve detection, road object localization, traffic scene analysis), decision making, end-to-end controlling and prediction, path and motion planning and augmented reality-based HUD, analyzing research works from 2011 to 2021 that focus on RGB camera vision. The literature is also analyzed for final representative outcomes as visualization in augmented reality-based head-up display (AR-HUD) with categories such as early warning, road markings for improved navigation and enhanced safety with overlapping on vehicles and pedestrians in extreme visual conditions to reduce collisions. The contribution of the literature review includes detailed analysis of current state-of-the-art deep learning methods that only rely on RGB camera vision rather than complex sensor fusion. It is expected to offer a pathway for the rapid development of cost-efficient and more secure practical autonomous vehicle systems.

Recent computer vision and deep learning breakthroughs have improved road safety by automatically classifying traffic signs. This research uses CNNs to classify traffic signs to improve road safety. Autonomous vehicles and intelligent driver assistance systems require accurate traffic sign detection and classification. Using deep learning, we created a CNN model that can recognize and classify road traffic signs. This research uses a massive dataset of labeled traffic sign photos for training and validation. These CNN algorithms evaluate images and produce real-time predictions to assist drivers and driverless cars in understanding traffic signs. Advanced driver assistance systems, navigation systems, and driverless vehicles can use this technology to give drivers more precise information, improving their decision-making and road safety. Researcher optimized CNN model design, training, and evaluation metrics during development. The model was rigorously tested and validated for robustness and classification accuracy. The research also solves real-world driving obstacles like illumination, weather, and traffic signal obstructions. This research shows deep learning-based traffic sign classification can dramatically improve road safety. This technology can prevent accidents and enhance traffic management by accurately recognizing and interpreting traffic signs. It is also a potential step toward a safer, more efficient transportation system with several automotive and intelligent transportation applications. Road safety is a global issue, and CNN-based traffic sign classification can reduce accidents and improve driving. On filter 3, Convolutional Neural Network training accuracy reached 98.9%, while validation accuracy reached 88.23%.

Understanding road scenes is crucial to the safe driving of autonomous vehicles, and object detection in road scenes is necessary to develop driving assistance systems. Current object detection algorithms are not very good at handling complex road scenes, and public datasets do not always adequately represent city traffic. Using Improved Multi-Scale Adaptive Feature Fusion Technology (IMSAFFT), this work suggests a real-time traffic information identification method to fix the issues of low detection accuracy of road scenes and high false detection rates in panoramic video images. In addition, a semantic recognition algorithm for a road scene based on image data is suggested. This study introduces computer vision-based approaches, including colour and texture recognition, object detection, and scene context understanding using Deep Neural Networks (DNN). An increasing number of deeper stacked layers allows the deep neural network to learn more complicated high-level semantic features, and the features' quality improves with time. A learning rate adaptive adjustment technique has been utilized to make training more efficient. After that, this improved detector is used to identify vehicles in original road environments. The suggested technique surpassed traditional detectors in the experiments with a high accuracy rate and processing speed. It worked well in real-world traffic situations for detecting overlapping, multiple, distant, and small objects. The simulation outcomes illustrate that the recommended IMSAFFT model increases the accuracy ratio of 98.4%, target image detection ratio of 97.4%, traffic prediction rate of 96.5%, processing speed rate of 10.4% and F1-score ratio of 95.4% compared to other existing models.

<span lang="EN-US">Intelligent transportation systems (ITS) are among the most focused research in this century. Actually, autonomous driving provides very advanced tasks in terms of road safety monitoring which include identifying dangers on the road and protecting pedestrians. In the last few years, deep learning (DL) approaches and especially convolutional neural networks (CNNs) have been extensively used to solve ITS problems such as traffic scene semantic segmentation and traffic signs classification. Semantic segmentation is an important task that has been addressed in computer vision (CV). Indeed, traffic scene semantic segmentation using CNNs requires high precision with few computational resources to perceive and segment the scene in real-time. However, we often find related work focusing only on one aspect, the precision, or the number of computational parameters. In this regard, we propose RBANet, a robust and lightweight CNN which uses a new proposed balanced attention module, and a new proposed residual module. Afterward, we have simulated our proposed RBANet using three loss functions to get the best combination using only 0.74M parameters. The RBANet has been evaluated on CamVid, the most used dataset in semantic segmentation, and it has performed well in terms of parameters’ requirements and precision compared to related work.</span>