Publications

Autonomous Vehicles (AVs) must interact with other road users. They must understand and adapt to complex pedestrian behaviour, especially during crossings where priority is not clearly defined. This includes feedback effects such as modelling a pedestrian’s likely behaviours resulting from changes in the AVs behaviour. For example, whether a pedestrian will yield if the AV accelerates, and vice versa. To enable such automated interactions, it is necessary for the AV to possess a statistical model of the pedestrian’s responses to its own actions. A previous work demonstrated a proof-ofconcept method to fit parameters to a simplified model based on data from a highly artificial discrete laboratory task with human subjects. The method was based on LIDAR-based person tracking, game theory, and Gaussian process analysis. The present study extends this method to enable analysis of more realistic continuous human experimental data. It shows for the first time how game-theoretic predictive parameters can be fit into pedestrians natural and continuous motion during road-crossings, and how predictions can be made about their interactions with AV controllers in similar real-world settings.

During the last 40 years, a large number of musical works was composed and recorded at the Centro di Sonologia Computazionale (CSC) of the University of Padova. The problem of how to preserve this increasing amount of audio documents arose and, in order to meet this need, the CSC started to carry out a research to develop a scientific methodology for preservation. In the last 7 years, this methodology was refined and applied to the digitization of more than 2,500 audio documents both from Italian and international audio archives, mainly stored on analogue magnetic carriers such as open-reel tapes and cassettes. Their content consists of electronic, folk and other kinds of music, as well as speech recordings. The methodology provides for collecting numerous metadata about the original carrier on which the audio information was stored, i.e. flange diameter, brand and material, its physical state and the related preservation copy, necessary for a correct preservation of the content and the contextual information. This work aims to extract and interpret the information from this wealth of data and metadata, which was collected and structured through the software PSKit. This may be useful for audio technicians, archives and policy makers to plan future preservation projects.

Providing technologies to support the visiting experience in cultural venues of artistic value is an important issue that needs to be addressed by considering the delicate nature of the places. Architectural heritage and visual arts are two valuable examples: the most sensible choice for augmenting the comprehension and the experience concerning this kind of cultural heritage is through audio cues (e.g., using audio guides). This work describes a pilot experiment to evaluate the impact of soundscapes and environmental acoustics reconstruction in the visitors’ experience. We proposed, to the visitors, an audioguide integrated with an optional experience presenting a comparison between a choir recorded in an anechoic chamber and the same choir as it would have sounded in the Church of the San Martino Charterhouse, using a preliminary acoustic survey. Experiments were conducted with real visitors to maximise the ecology of the collected data and results show that people are significantly interested towards the proposed experience, motivating further efforts to improve the quality and depth of the contents provided with this strategy.

Objectives: The purpose of this study was to assess the functional and prognostic correlates of B-lines during stress echocardiography (SE). Background: B-profile detected by lung ultrasound (LUS) is a sign of pulmonary congestion during SE. Methods: The authors prospectively performed transthoracic echocardiography (TTE) and LUS in 2,145 patients referred for exercise (n = 1,012), vasodilator (n = 1,054), or dobutamine (n = 79) SE in 11 certified centers. B-lines were evaluated in a 4-site simplified scan (each site scored from 0: A-lines to 10: white lung for coalescing B-lines). During stress the following were also analyzed: stress-induced new regional wall motion abnormalities in 2 contiguous segments; reduced left ventricular contractile reserve (peak/rest based on force, ≤2.0 for exercise and dobutamine, ≤1.1 for vasodilators); and abnormal coronary flow velocity reserve ≤2.0, assessed by pulsed-wave Doppler sampling in left anterior descending coronary artery and abnormal heart rate reserve (peak/rest heart rate) ≤1.80 for exercise and dobutamine (≤1.22 for vasodilators). All patients completed follow-up. Results: According to B-lines at peak stress patients were divided into 4 different groups: group I, absence of stress B-lines (score: 0 to 1; n = 1,389; 64.7%); group II, mild B-lines (score: 2 to 4; n = 428; 20%); group III, moderate B-lines (score: 5 to 9; n = 209; 9.7%) and group IV, severe B-lines (score: ≥10; n = 119; 5.4%). During median follow-up of 15.2 months (interquartile range: 12 to 20 months) there were 38 deaths and 28 nonfatal myocardial infarctions in 64 patients. At multivariable analysis, severe stress B-lines (hazard ratio [HR]: 3.544; 95% confidence interval [CI]: 1.466 to 8.687; p = 0.006), abnormal heart rate reserve (HR: 2.276; 95% CI: 1.215 to 4.262; p = 0.010), abnormal coronary flow velocity reserve (HR: 2.178; 95% CI: 1.059 to 4.479; p = 0.034), and age (HR: 1.031; 95% CI: 1.002 to 1.062; p = 0.037) were independent predictors of death and nonfatal myocardial infarction. Conclusions: Severe stress B-lines predict death and nonfatal myocardial infarction. (Stress Echo 2020–The International Stress Echo Study [SE2020]; NCT03049995)

Background: The efficacy and effectiveness of nintedanib as a first-line therapy in idiopathic pulmonary fibrosis (IPF) patients have been demonstrated by clinical trials and real-life studies. Our aim was to examine the safety profile and effectiveness of nintedanib when it is utilized as a second-line treatment in subjects who have discontinued pirfenidone. Methods: The medical charts of 12 patients who were switched from pirfenidone to nintedanib were examined retrospectively. The drug’s safety was defined by the number of adverse events (AEs) that were reported; disease progression was evaluated based on the patient’s vital status and changes in forced vital capacity (FVC) at 12-month follow-up. Results: The numbers of patients experiencing AEs and of the AEs per patient in our study group didn’t significantly differ with respect to a group of 56 individuals who were taking nintedanib as a first-line therapy during the study period (5/12 vs. 22/56; p = 0.9999, and 0.00 (0.00–1.00) vs. 0.00 (0.00–3.00); p = 0.517, respectively). Two out of the 3 patients who had been switched to nintedanib due to a rapid disease progression showed stabilized FVC values. Conclusions: Nintedanib was found to have an acceptable safety profile in the majority of the IPF patients switched from pirfenidone. Prospective studies are warranted to determine if the drug can effectively delay disease progression in these patients.

Approximating the stationary probability of a state in a Markov chain through Markov chain Monte Carlo techniques is, in general, inefficient. Standard random walk approaches require Õ(τ/ π(v)) operations to approximate the probability π(v) of a state v in a chain with mixing time τ, and even the best available techniques still have complexity Õ(τ1.5/ π(v) 0.5) ; and since these complexities depend inversely on π(v), they can grow beyond any bound in the size of the chain or in its mixing time. In this paper we show that, for time-reversible Markov chains, there exists a simple randomized approximation algorithm that breaks this “small-π(v) barrier”.

A survey of fresh great Mediterranean scallops (Pecten jacobaeus) and frozen queen scallops (Aequipecten opercularis) collected from the northern Adriatic Sea demonstrated the presence of nematode larvae that were morphologically and molecularly identified as Sulcascaris sulcata. The presence of this nematode in the northern Adriatic Sea has been described in the loggerhead sea turtle, which is the final host, but it was only recently described in the intermediate molluscan host. Affected scallops present rust-brown lesions in the adductor muscle. In this study, necroscopic and histopathological lesions are described in affected scallops. Furthermore, the detrimental effects of this parasite on the commercialization of scallops under the current European legislation concerning the hygiene of seafood products are evaluated.

Farming robots need a fast and robust image segmentation module to apply targeted treatments, which require the ability to distinguish, in real time, between crop and weeds. Existing solutions make use of visual classifiers that are trained on large annotated datasets. However, generating large datasets with pixel-wise annotations is an extremely time-consuming task. In this work, we tackle the crop/weed segmentation problem by using a synthetic image generation method to augment the training dataset without the need of manually labelling the images. The proposed approach consists in training a Generative Adversarial Network (GAN), which can automatically generate realistic agricultural scenes. As a difference with respect to common GAN approaches, where the network learns how to reproduce an entire scene, we generate only instances of the objects of interest in the scene, namely crops. This allows to build a generative model that is more compact and easier to train. The generated objects are then placed into real images of agricultural datasets, thus creating new images that can be used for training. To evaluate the performance of the proposed approach, quantitative experiments have been carried out using different segmentation network architectures, showing that our method well generalizes across multiple architectures.

Digital technology in museum practice provides new means of interaction with artifacts and collections. In particular, we need interactive installations in order to encourage and stimulate visitors to learn and understand archaeological musical instruments through engagement and active participation: these instruments (i.e., interactive artifacts per se) are de facto unplayable and inaccessible to visitors, as a consequence of their preservation issues. However, presenting artifacts to the general public is a complex task for their multifaceted nature, and digital technology must not sacrifice accuracy or depth of information for the sake of entertainment. Moreover, deploying digital technology is a multidisciplinary effort that requires an interplay among different fields, from history and archaeology to information engineering and craftsmanship. In this article, we present a methodology to relate such disciplines in order to design a digital multimedia installation that promotes archaeological musical instruments in a museum. In defining the problem, we identify four different aspects to consider: the museum collection, the museum environment, the manufacturing opportunities for the installation, and the user experience. Such aspects are integrated in a design approach that is centered on Design Thinking. The proposed methodology is exemplified in the designing and manufacturing of an installation for a Pan flute from Egypt dated back to 700 A.D., a case in which multisensory interaction is particularly important to convey the lost sound of the instrument. We describe in detail an installation (exhibited at the Museum of Archaeological Science and Art at the University of Padova), which virtually recreates the Pan flute and communicates information related to its history, iconography, acoustics, and musicology. Just after the deployment of the installation, we also carried out an assessment with a group of experts in the fields of information engineering, music, musicology, and archaeology. The good results obtained demonstrate that the installation is a convenient way of interaction, simple to use and aesthetically integrated in the museum context.

Archives often include documents that can hardly be considered publications or grey literature as such, yet they maintain their documentary value and play a role of primary sources for the specialists. These documents, indeed, can help archivists to reveal the sedimentation process of the archive itself and to preserve the authentic context of the documentary production. They also appear to be very useful for the community of researchers and scholars. This happens more frequently with oral archives which include ‘non-conventional sources’, thus bringing together audio documents, fieldworks notes, correspondence, slipcases, analogic compact cassettes or open reels. At the cross-road of two disciplines, Archival Science and Grey Literature, this paper aims to argue the applicability of the concept of grey literature to this wide range of documentary materials, by showing the experience of Archivio Vi.Vo, a regional project aiming at building a model for archiving, preserving, managing and disseminating audio documents.