Advanced search in
Research products
arrow_drop_down
Searching FieldsTerms
Any field
arrow_drop_down
includes
arrow_drop_down
Include:
138 Research products, page 1 of 14

  • Publications
  • Research data
  • Research software
  • Conference object
  • Part of book or chapter of book
  • 0202 electrical engineering
  • Flore (Florence Research Repository)

10
arrow_drop_down
Relevance
arrow_drop_down
  • Publication . Part of book or chapter of book . Conference object . 2014
    Open Access
    Authors: 
    Duccio Troiano; Andrés García Morro; Alessandro Merlo; Eduardo Vendrell Vidal;
    Publisher: Springer International Publishing
    Country: Italy

    Despite extensive research having been conducted on the subject, the problem of three-dimensional information systems for historical cities is actually still unresolved. In addition, commercially available software seems to be increasingly aiming at a quick development of unspecific urban settings, rather than at a metrically and perceptively faithful representation of reality. In this scenario, the SIUR 3D software (Sistema Informativo URbano tridimensionale) is based on a management structure that links an interactive, photorealistic and metrically reliable model of a city with a qualitative database of the historical, archaeological and material scope of an architectural part. Such application uses the Unity 3D game engine for geometrical models management and is equipped for online data sharing.

  • Publication . Conference object . 2017
    Open Access
    Authors: 
    Tommaso Zoppi; Andrea Ceccarelli; Andrea Bondavalli;
    Publisher: ACM
    Country: Italy

    The loosely coupled interoperability of heterogeneous existing systems, together with the ongoing replacement of monolithic systems design with Off-The-Shelf (OTS) approaches, promotes a new architectural paradigm that is called System of Systems (SoS). In SoSs, independent and autonomous constituent systems (CSs) cooperate to achieve higher-level goals. Some inherent challenges are that boundaries of the SoS may be partially unknown and the components may be governed by different authorities, affecting the ability to observe the system as a whole. Further, novel challenges related to dependability and security are introduced, such as the detection of emerging and possibly unexpected behaviors resulting from the interconnection of previous disconnected CSs. In this paper we explore these challenges questioning if a novel mindset to error, malware or intrusion detection is needed when dealing with SoSs. With the support of a state of the art review, we first identify the design principles and the performance targets of a monitoring and anomaly detection framework. Then we discuss these principles at the light of SoS fundamentals. Ultimately, we propose an approach to design a monitoring and anomaly detection framework for SoSs aggregating i) monitoring approaches ii) SoS properties, and iii) anomaly detection techniques.

  • Publication . Conference object . 2017
    Open Access
    Authors: 
    Claudio Badii; Pierfrancesco Bellini; Daniele Cenni; Angelo Difino; Michela Paolucci; Paolo Nesi;
    Publisher: Institute of Electrical and Electronics Engineers Inc.
    Country: Italy

    The new challenges in the smart city context are mainly related to the stimulation of the city users towards taking more sustainable behaviors, in mobility and energy. The state of the art in this case is mainly focused on classical smart city solution for informing the city users and or for engaging them with specific wired rules toward virtuous models. And not using flexible languages and predictive models, pushing them towards a larger range of virtuous habits. On this regards, the main problems are the computation of user behavior via data analytic (semantic computing, machine learning), as well as the formalization of strategies via simple and well formalized language for producing engagements to the city users, which can be understood by city operators. In this paper, a solution for city users engagement is studied and implemented for Sii-Mobility Smart city national project in Italy has been presented. The solution has been implemented thanks to the exploitation of Km4City model and semantic computing. The paper also presents the validation of results about the effective usage of the solution by providing some statistical evidence about the efficient assessment of user behavior and of engagement rules acceptance rate.

  • Publication . Conference object . Preprint . Article . 2020
    Open Access
    Authors: 
    Federico Vaccaro; Marco Bertini; Tiberio Uricchio; Alberto Del Bimbo;
    Publisher: ACM
    Country: Italy

    In this paper, we address the problem of image retrieval by learning images representation based on the activations of a Convolutional Neural Network. We present an end-to-end trainable network architecture that exploits a novel multi-scale local pooling based on NetVLAD and a triplet mining procedure based on samples difficulty to obtain an effective image representation. Extensive experiments show that our approach is able to reach state-of-the-art results on three standard datasets. Comment: Accepted at ICMR 2020

  • Publication . Conference object . 2018
    Open Access
    Authors: 
    Francesco Gelli; Tiberio Uricchio; Xiangnan He; Alberto Del Bimbo; Tat-Seng Chua;
    Publisher: ACM
    Country: Italy

    Brands and organizations are using social networks such as Instagram to share image or video posts regularly, in order to engage and maximize their presence to the users. Differently from the traditional advertising paradigm, these posts feature not only specific products, but also the value and philosophy of the brand, known as brand associations in marketing literature. In fact, marketers are spending considerable resources to generate their content in-house, and increasingly often, to discover and repost the content generated by users. However, to choose the right posts for a brand in social media remains an open problem. Driven by this real-life application, we define the new task of content discovery for brands, which aims to discover posts that match the marketing value and brand associations of a target brand. We identify two main challenges in this new task: high inter-brand similarity and brand-post sparsity; and propose a tailored content-based learning-to-rank system to discover content for a target brand. Specifically, our method learns fine-grained brand representation via explicit modeling of brand associations, which can be interpreted as visual words shared among brands. We collected a new large-scale Instagram dataset, consisting of more than 1.1 million image and video posts from the history of 927 brands of fourteen verticals such as food and fashion. Extensive experiments indicate that our model can effectively learn fine-grained brand representations and outperform the closest state-of-the-art solutions.

  • Open Access
    Authors: 
    Mohamad Gharib; Paolo Lollini; Andrea Bondavalli;
    Publisher: Institute of Electrical and Electronics Engineers Inc.
    Country: Italy
    Project: EC | DEVASSES (612569)

    A System-of-Systems (SoS) is an integration of a finite number of Constituent Systems (CSs), which are networked together for achieving a certain higher goal. Therefore, integration is the key viability of any SoS. Although the integration of CSs can be achieved by the exchange of information, no existing work has considered the quality of such information. Without considering Information Quality (IQ), a CS may depend on inaccurate, incomplete, inconsistent, invalid, and/or untrustworthy information, which might lead to its failure, and in turn to catastrophic incidents in the case of critical SoS. The main objective of the paper is proposing a novel conceptual model that provides the required concepts for analyzing for SoS. We illustrate the utility of the model with an example concerning the Intelligent Transportation System (ITS) domain.

  • Open Access
    Authors: 
    Matteo Franchi; Alessandro Ridolfi; Leonardo Zacchini;
    Publisher: Elsevier B.V.
    Country: Italy
    Project: EC | EUMarineRobots (731103)

    Abstract This paper proposes an underwater navigation system where linear speed estimations, obtained with a 2D Forward-Looking SONAR (FLS), are integrated within a navigation filter and this solution is shown to work satisfyingly in the absence of Doppler Velocity Log (DVL) readings. Both to provide a better description of the system, which is a dynamic entity in a dynamic environment and to characterize FLS measurements, an Adaptive Unscented Kalman Filter (AUKF)-based estimator is here proposed. The solution has been tested and validated offline making use of navigation data obtained during sea trials performed in July 2018 with FeelHippo AUV at the basin of the NATO Science and Technology Organization Centre for Maritime Research and Experimentation (CMRE), La Spezia (Italy).

  • Publication . Conference object . Preprint . Article . 2021
    Open Access English
    Authors: 
    My Kieu; Lorenzo Berlincioni; Leonardo Galteri; Marco Bertini; Andrew D. Bagdanov; Alberto Del Bimbo;
    Country: Italy

    In this paper we propose a method for improving pedestrian detection in the thermal domain using two stages: first, a generative data augmentation approach is used, then a domain adaptation method using generated data adapts an RGB pedestrian detector. Our model, based on the Least-Squares Generative Adversarial Network, is trained to synthesize realistic thermal versions of input RGB images which are then used to augment the limited amount of labeled thermal pedestrian images available for training. We apply our generative data augmentation strategy in order to adapt a pretrained YOLOv3 pedestrian detector to detection in the thermal-only domain. Experimental results demonstrate the effectiveness of our approach: using less than 50\% of available real thermal training data, and relying on synthesized data generated by our model in the domain adaptation phase, our detector achieves state-of-the-art results on the KAIST Multispectral Pedestrian Detection Benchmark; even if more real thermal data is available adding GAN generated images to the training data results in improved performance, thus showing that these images act as an effective form of data augmentation. To the best of our knowledge, our detector achieves the best single-modality detection results on KAIST with respect to the state-of-the-art. Accepted at ICPR2020

  • Open Access
    Authors: 
    Edoardo Locorotondo; Vincenzo Cultrera; Luca Pugi; Lorenzo Berzi; M. Pasquali; Natascia Andrenacci; Giovanni Lutzemberger; Marco Pierini;
    Publisher: IEEE
    Country: Italy
    Project: EC | OBELICS (769506)

    The aging behavior of lithium cell has a profound impact on its performance in terms of energy and power efficiency, especially when it is considered in End Of Life (EOL) in automotive field. Lithium battery is considered in EOL if at 85-80% of nominal capacity. Today, the reusing of Electric and Hybrid Vehicles EOL batteries on stationary applications, giving a second life to these batteries, is a solution to reduce high potential cost of lithium batteries. Currently, there is a lack of investigation of the performances of these second life batteries. This paper depicts the performance results of five NMC cells at different SOH, where four of these cells are considered in EOL, so ready to be investigated for possible second use. By results, there are many way to correlate battery SOH and battery performance, e.g. an increase of the internal resistance and the constant-voltage (CV) phase charging duration, the change of the open circuit voltage shape curve. Finally, a battery model based on electrical equivalent circuit is build and implemented in Matlab/Simulink, which is validated by comparison between voltage experimental and simulated data.

  • Publication . Part of book or chapter of book . Conference object . 2018
    Open Access
    Authors: 
    Samuele Capobianco; Leonardo Scommegna; Simone Marinai;
    Publisher: Springer International Publishing
    Country: Italy

    In this work we propose one deep architecture to identify text and not-text regions in historical handwritten documents. In particular we adopt the U-net architecture in combination with a suitable weighted loss function in order to put more emphasis on most critical areas. We define one weighted map to balance the pixel frequency among classes and to guide the training with local prior rules. In the experiments we evaluate the performance of the U-net architecture and of the weighted training on one benchmark dataset. We obtain good results using global metrics improving global and local classification scores.

Send a message
How can we help?
We usually respond in a few hours.