Asphalt pavements are subject to regular inspection and maintenance activities over time. Many techniques have been suggested to evaluate pavement surface conditions, but most of these are either labour-intensive tasks or require costly instruments. This article describes a robust intelligent pavement distress inspection system that uses cost-effective equipment and the ‘you only look once’ detection algorithm (YOLOv3). A dataset for flexible pavement distress detection with around 13,135 images and 30,989 bounding boxes of damage was used during the neural network training, calibration, and validation phases. During the testing phase, the model achieved a mean average precision of up to 80%, depending on the type of pavement distress. The performance metrics (loss, precision, recall, and RMSE) that were applied to estimate the object detection accuracy demonstrate that the technique can distinguish between different types of asphalt pavement damage with remarkable accuracy and precision. Moreover, the confusion matrix obtained in the validation process shows a distress classification sensitivity of up to 98.7%. The suggested technique was successfully implemented in an inspection car. Measurements conducted on urban roads crossed by tramway lines in the city of Palermo proved the real-time ability and great efficacy of the detection system, with potentially remarkable advances in asphalt pavement examination efficacy due to the high rates of correct distress detection.

Artificial intelligence and deep learning-based techniques undoubtedly are the future of Advanced Driver-Assistance Systems (ADAS) technologies. In this article is presented a technique for detecting, recognizing and tracking pedestrians, vehicles and cyclists along a tramway infrastructure in a complex urban environment by Computer Vision, Deep Learning approaches and YOLOv3 algorithm. Experiments have been conducted in the tramway Line 2 “Borgonuovo –Notarbartolo” (Palermo, Italy) in correspondence of the tramway segments crossing a roundabout having an external diameter of 24 m. A survey vehicle equipped with a video camera was used in the study. The results of the research show that the proposed method is able to search and detect the position and the speed of road users near and over the rails in front of the tram in a very precise way as demonstrate by the estimated values of the Accuracy, Loss and Precision obtained during the neural networks training process. The implementation of this advanced detection method in ADAS systems may increase the safety of novel autonomous trams and autonomous rapid trams (ARTs).

Background: Hepatic encephalopathy (HE) may arise as a possible consequence of cirrhosis. Magnetic resonance imaging (MRI) may reveal a T1-weighted hyperintensity in the globi pallidi, indicating the deposition of paramagnetic substances. The objective of this research was to implement a machine learning-based radiomic model to predict the diagnosis and severity of chronic hepatic encephalopathy in adult patients with cirrhosis. Methods: Between October 2018 and February 2020, brain magnetic resonance imaging (MRI) was conducted on adult patients, both with and without cirrhosis. The control population consisted of individuals who did not have a previous medical record of chronic liver disease. The grade of hepatic encephalopathy (HE) was determined by considering factors such as the presence of underlying liver disease, the severity of clinical symptoms, and the frequency of encephalopathic episodes. Radiomic texture analysis based on five machine learning algorithms was applied to axial T1-weighted MR images of bilateral lentiform nuclei. Using the area under the receiver operating characteristics curve, we determined the accuracy of the five machine learning-based algorithms in predicting the presence of HE and the HE grading. Results: The ultimate research cohort included 124 individuals, with 70 being cirrhotic patients and 54 being non-cirrhotic controls. Of the total number of patients, 38 had a previous occurrence of HE and, among them, 22 had a grade of HE greater than 1. The multilayer perceptron algorithm classified patients versus controls with an accuracy of 100%. The k-nearest neighbor (KNN) algorithm classified patients with or without HE with an accuracy of 76.5%. The multilayer perceptron algorithm classified HE grade (HE grade 1, HE grade ≥ 2) with an accuracy of 94.1%. Conclusions: The machine learning algorithms implemented provide a robust modeling technique for deriving valuable insights from brain MR images in cirrhotic patients and this can serve as an imaging tool valuable for the assessment of the burden of hepatic encephalopathy.

This research put forward an efficacious real-time deep learning-based technique to automate the process of monitoring the social distancing in transportation systems (e.g., bus stops, railway stations, airport terminals, etc.) and other public spaces with the purpose to mitigate the impact of coronavirus pandemic. The proposed technique makes use of the YOLOv3 model to segregate humans from the background of each image of a surveillance video and the linear Kalman filter for tracking the humans’ motion even in case in which another object or person overlaps the trajectory of the person under analysis. The performance of the model in human detection is extremely high as demonstrated by the accuracy of the model that reaches values higher than 95%. The detection algorithm can be applied for alerting people to keep a safe distance from each other when they are in crowded places or in groups.