Search

Abstract Intersection markings play a vital role in providing road users with guidance and information. The conditions of intersection markings will be gradually degrading due to vehicular traffic, rain, and/or snowplowing. Degraded markings can confuse drivers, leading to increased risk of traffic crashes. Timely obtaining high-quality information of intersection markings lays a foundation for making informed decisions in safety management and maintenance prioritization. However, current labor-intensive and high-cost data collection practices make it very challenging to gather intersection data on a large scale. This paper develops an automated system to intelligently detect intersection markings and to assess their degradation conditions with existing roadway Geographic information systems (GIS) data and aerial images. The system harnesses emerging artificial intelligence (AI) techniques such as deep learning and multi-task learning to enhance its robustness, accuracy, and computational efficiency. AI models were developed to detect lane-use arrows (85% mean average precision) and crosswalks (89% mean average precision) and to assess the degradation conditions of markings (91% overall accuracy for lane-use arrows and 83% for crosswalks). Data acquisition and computer vision modules developed were integrated and a graphical user interface (GUI) was built for the system. The proposed system can fully automate the processes of marking data collection and condition assessment on a large scale with almost zero cost and short processing time. The developed system has great potential to propel urban science forward by providing fundamental urban infrastructure data for analysis and decision-making across various critical areas such as data-driven safety management and prioritization of infrastructure maintenance.

Deep learning-based intelligent vehicle perception has been developing prominently in recent years to provide a reliable source for motion planning and decision making in autonomous driving. A large number of powerful deep learning-based methods can achieve excellent performance in solving various perception problems of autonomous driving. However, these deep learning methods still have several limitations, for example, the assumption that lab-training (source domain) and real-testing (target domain) data follow the same feature distribution may not be practical in the real world. There is often a dramatic domain gap between them in many real-world cases. As a solution to this challenge, deep transfer learning can handle situations excellently by transferring the knowledge from one domain to another. Deep transfer learning aims to improve task performance in a new domain by leveraging the knowledge of similar tasks learned in another domain before. Nevertheless, there are currently no survey papers on the topic of deep transfer learning for intelligent vehicle perception. To the best of our knowledge, this paper represents the first comprehensive survey on the topic of the deep transfer learning for intelligent vehicle perception. This paper discusses the domain gaps related to the differences of sensor, data, and model for the intelligent vehicle perception. The recent applications, challenges, future researches in intelligent vehicle perception are also explored.

As one of the fundamental tasks of autonomous driving, depth perception aims to perceive physical objects in three dimensions and to judge their distances away from the ego vehicle. Although great efforts have been made for depth perception, LiDAR-based and camera-based solutions have limitations with low accuracy and poor robustness for noise input. With the integration of monocular cameras and LiDAR sensors in autonomous vehicles, in this article, we introduce a two-stream architecture to learn the modality interaction representation under the guidance of an image reconstruction task to compensate for the deficiencies of each modality in a parallel manner. Specifically, in the two-stream architecture, the multi-scale cross-modality interactions are preserved via a cascading interaction network under the guidance of the reconstruction task. Next, the shared representation of modality interaction is integrated to infer the dense depth map due to the complementarity and heterogeneity of the two modalities. We evaluated the proposed solution on the KITTI dataset and CALAR synthetic dataset. Our experimental results show that learning the coupled interaction of modalities under the guidance of an auxiliary task can lead to significant performance improvements. Furthermore, our approach is competitive against the state-of-the-art models and robust against the noisy input. The source code is available at https://github.com/tonyFengye/Code/tree/master. [ABSTRACT FROM AUTHOR]

Abstract The distracted phone‐use behaviours among pedestrians, like Texting, Game Playing and Phone Calls, have caused increasing fatalities and injuries. However, the research of phone‐related distracted behaviour by pedestrians has not been systemically studied. It is desired to improve both the driving and pedestrian safety by automatically discovering the phone‐related pedestrian distracted behaviours. Herein, a new computer vision‐based method is proposed to detect the phone‐related pedestrian distracted behaviours from a view of intelligent and autonomous driving. Specifically, the first end‐to‐end deep learning based Two‐Branch Convolutional Neural Network (CNN) is designed for this task. Taking one synchronised image pair by two front on‐car GoPro cameras as the inputs, the proposed two‐branch CNN will extract features for each camera, fuse the extracted features and perform a robust classification. This method can also be easily extended to video‐based classification by confidence accumulation and voting. A new benchmark dataset of 448 synchronised video pairs of 53,760 images collected on a vehicle is proposed for this research. The experimental results show that using two synchronised cameras obtained better performance than using one single camera. Finally, the proposed method achieved an overall best classification accuracy of 84.3% on the new benchmark when compared to other methods.

ObjectivesThe purpose of this study was to evaluate the feasibility and diagnostic performance of prostate-specific membrane antigen (PSMA) based 18F-DCFPyL PET/CT-ultrasound (PET/CT-US) or PET/MRI-ultrasound (PET/MRI-US) fusion targeted biopsy for intra-prostatic PET-positive lesions.MethodsFrom April 2018 to November 2019, we prospectively enrolled 55 candidates to perform PET/CT-US or PET/MRI-US fusion targeted biopsies for solitary PET-positive prostate lesions (two to four cores/lesion). The positive rates of prostate cancer based on patients and biopsy cores were calculated respectively. With reference to the pathological results of biopsy cores, the MR signal characteristics in the area of the PET-positive lesion were analyzed for the patients who underwent PET/MRI.ResultsA total of 178 biopsy cores were taken on the 55 patients. One hundred forty-six biopsy cores (82.0%, 146/178) from 51 (92.7%, 51/55) patients were positive for prostate cancer; 47 (85.5%, 47/55) were clinically significant prostate cancer. It is noteworthy that nine patients underwent both 18F-DCFPyL PET/CT and PET/MRI examinations; the seven patients with prostate cancer showed abnormal MR signal in the area of the PET-positive lesion while the other two patients with prostatic hyperplasia and prostatitis showed normal MR signal in the area of the PET-positive lesion.ConclusionThis study indicated that 18F-DCFPyL PET/CT-US or PET/MRI-US fusion targeted prostate biopsies may be valuable for prostate cancer diagnosis and have a high detection rate of clinically significant prostate cancer for PET-positive lesions. PET/MR can rule out some false PET-positive lesions, which may potentially reduce unnecessary prostate biopsies.