130,555 research outputs found
Les premiers fédéralistes bulgares
No full textMeneghello-Dincic Kruno. Les premiers fédéralistes bulgares. In: Revue d’histoire moderne et contemporaine, tome 5 N°4, Octobre-d�cembre 1958. pp. 295-313
A volumetric 3-D digital analysis of dimensional changes to the alveolar process at implants placed immediately into extraction sockets
No full textThe objective of this study was to validate the use of a novel method to elaborate 3-D data on dimensional changes to the alveolar process after one year of healing at implants placed immediately into extraction sockets
Mobility prediction via sequential learning for 5G mobile networks
No full textHere, we present a mobility prediction framework for 5G mobile systems. Our work stems from the intuition that mobility in vehicular networks is highly correlated, and such correlation can be captured by advanced neural network designs to anticipate the users' point of attachment. To prove this, we combine Markov chains with recurrent and convolutional neural networks, training them on mobility trajectories estimated by the received radio signal from mobile millimeter-wave devices. The proposed framework is decentralized, i.e., user trajectories are independently learned by each base station. In this paper, various problems are pragmatically tackled and solved, such as dealing with imbalanced datasets, as some trajectories are under represented, and obtaining a mobility classifier whose accuracy increases as new mobility samples are collected.The proposed technique is assessed using emulated traces obtained through the SUMO mobility simulator for the city of Cologne. Numerical results show accuracies higher than 88% in the prediction of the next serving base station from 4 seconds before the handover is performed. Mobility (next base station) predictors like the ones presented here are key for network management purposes within 5G networks, e.g., to proactively allocate communication and edge computing resources
HIJACK: Learning-based Strategies for Sound Classification Robustness to Adversarial Noise
No full textThe effective deployment of smart service systems within homes, workspaces and cities, requires gaining context and situational awareness to take action when changes are detected. To this end, sound classification systems are widely adopted and integrated into several smart devices to continuously monitor the environment. However, sound classification algorithms are prone to adversarial attacks that pose a considerable security threat to smart service systems where they are integrated. In this paper, we devise HIJACK, a novel machine learning framework entailing five neural network strategies to enforce the robustness of sound classification systems to adversarial noise injection. The HIJACK methodologies can be applied to any neural network-based sound classifier and consist of tailored transformations of the input audio during training along with specific additional layers added to the neural network architecture. To assess the noise robustness provided by the HIJACK strategies, we design a measure based on a L-2-adversarial attack to sound classification identified as the normalized fast gradient method (NFGM) - that constructs the adversarial noise by maximizing the sound mis-classification probability. We assessed the robustness of HIJACK to the proposed NFGM attack on a publicly available dataset. The results show that the combination of the five HIJACK strategies allows reaching robustness to adversarial noise 58 times larger than state-of-the-art neural networks for sound classification, guaranteeing a classification accuracy above 83%
3-D laser scan reconstruction of the face and jaws: a new approach in cranio-maxillofacial surgery
No full textThe aim of this study was to finely adjust a method for the construction of a 3D-textured digital model of the face and dental arches of the maxilla and mandible from a laser scan of the face and plaster casts of the mouth and dental arches of patients with craniofacial malformations, and reconstruction after facial trauma or resection for neoplasm of facial skeleton. This is motivated by the need to obtain a virtual copy of the patient’s face at no biological cost as an alternative to the method which involves using CT. The method based on CT provides a great deal of information, but the radiation absorbed by the patient is very high.
Materials and methods
The 3-D Laser scanner Konica Minolta VIVID 910 is used to scan the surface of the patient’s face from different perspectives. These scanning perspectives are then recombined to reconstruct the three-dimensional facial features with minimal error in order to reconstruct the 3D-textured model of the face. A bite fork adapted to the purpose is used as the 3-D reference mark for repositioning the scan of the maxillar and mandibular arch plaster casts within the face model. The acquisition method was validated by verifying the consistency of several anthropometric measurements recorded in vivo with the corresponding measurements recorded on the virtual computer model.
Results
The patient’s virtual face reconstructed by means of meshes can, be analysed as it is or can be later on converted into a NURBS (Nonuniform Rational B-Splines) surface. This virtual model is used not only to extrapolate isoparametric curves reproducing facial sections in the three spatial planes, but also to obtain further metric information regarding surface and volumes. This enables the patient’s profile to be studied in the usual way, but also provides other useful anthropometric features, its volumetric characteristics and the changes induced by cranio-maxillofacial surgery.
Conclusions
The facial skin surface may serve as a sufficiently stable and invariable entity for registering patients for computer-assisted cranio-maxillofacial surgery. The proposed method proves an excellent way to study all the morphological facial features of patients with cranio-facial malformation or after facial trauma or resection for neoplasm of facial skeleton. The collection of an appropriate database of case studies, before and after surgical treatment, will enable us to implement a software for correlating dental arch displacements with changes in the corresponding facial surface
Ippolito Nievo e Luigi Meneghello: una lettura stratificata
No full textThe essay focuses on the relationship between Luigi Meneghello’s works and Le Confessioni d’un Italiano by Ippolito Nievo. Although several scholars have found similarities between the nineteenth century masterpiece and Meneghello’s works, there aren’t clear references to Nievo in his writings, except for just a mention in Confessioni. The research develops through three points of view: firstly, the investigation of Meneghello’s educational and cultural profile in regard to 19th century; secondly, an overview of correspondences between Confessioni and Meneghello’s works (such as the purpose of a truthful, anti-heroic testimony through irony), ending with a relevant example of similar scenes; thirdly a part dedicated to the edition of Confessioni from Meneghello’s library. The two volumes are important because of the year of publication, 1960, and also for annotations Meneghello wrote along their margins
A 3-Dimensional Facial Morpho-Dynamic Database in the development of a Prediction Model in Orthognathic Surgery
No full textCurrent methodologies in the prevision of post-surgical features of the face in orthognathic surgery are mainly 2-D. An improvement is certainly given by the introduction of CT, but its acceptance is controversial due to its high biological cost. As an alternative, in this study an effective procedure for the construction of a 3-D textured digital model of the face and dental arches of patients with dentofacial malformations using a 3-D laser scanner at no biological cost is presented. A 3-D Laser scanner Konica-Minolta VIVID 910 is used to obtain multiple scans from different perspectives of the face of patients with dentofacial malocclusions requiring orthognathic surgery. These multiple views are then recombined, integrating also the maxillary and mandibular arch plaster casts, to obtain the 3-D textured model of the face and occlusion with minimal error. A viable methodology was identified for the face and occlusal modeling of orthognathic patients and validated in a test case, confirming its effectiveness: the 3-D model created accurately describes the actual features of the patient's face; the proposed methodology can be easily applied in the clinical routine to accurately record the steps of the surgical treatment and to perform accurate anthropometric analyses of the facial morphology, and thus constitute the necessary database for the development of previsional tools in orthognathic surgery. The proposed method is effective in recording all the morphological facial features of patients with dentofacial malformations, to develop a facial modification database and tools for virtual surgery
RIMS_tool: a Hybrid Simulator for Business Processes
No full textBusiness Process Simulation represents a powerful instrument for business analysts when analyzing and
comparing business processes. Most of the state-of-the-art business process simulators, however, rely on
Discrete event simulation, which requires various unrealistic assumptions and simplifications to perform
experiments. Predictive Process Monitoring, on the other hand, offers a viable way to complete ongoing
traces or to generate entire traces from scratch, via predictions of the next activities and their attributes.
Predictive models, though, are usually based on black-box approaches that make it difficult to reason
on what-if scenarios. RIMS_Tool is a hybrid business process simulator that aims at combining predictive
models built from data and Discrete event simulation at runtime in a white-box manner. The proposed
tool, thus, is able to exploit the strengths and avoid the limitations of both approaches
MeSH term explosion and author rank improve expert recommendations
No full textInformation overload is an often-cited phenomenon that reduces the productivity, efficiency and efficacy of scientists. One challenge for scientists is to find appropriate collaborators in their research. The literature describes various solutions to the problem of expertise location, but most current approaches do not appear to be very suitable for expert recommendations in biomedical research. In this study, we present the development and initial evaluation of a vector space model-based algorithm to calculate researcher similarity using four inputs: 1) MeSH terms of publications; 2) MeSH terms and author rank; 3) exploded MeSH terms; and 4) exploded MeSH terms and author rank. We developed and evaluated the algorithm using a data set of 17,525 authors and their 22,542 papers. On average, our algorithms correctly predicted 2.5 of the top 5/10 coauthors of individual scientists. Exploded MeSH and author rank outperformed all other algorithms in accuracy, followed closely by MeSH and author rank. Our results show that the accuracy of MeSH term-based matching can be enhanced with other metadata such as author rank
- …
