Piero Fraternali

Piero Fraternali
Politecnico di Milano | Polimi · Department of Electronics, Information, and Bioengineering

PhD in Computer Science

About

323
Publications
108,001
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
7,973
Citations
Introduction
Working on human computation, games with a purpose and crowdsourcing.
Additional affiliations
November 1990 - present
Politecnico di Milano
Position
  • Professor (Full)

Publications

Publications (323)
Article
Full-text available
Non-neural machine learning (ML) and deep learning (DL) are used to predict system failures in industrial maintenance. However, only a few studies have assessed the effect of varying the amount of past data used to make a prediction and the extension in the future of the forecast. This study evaluates the impact of the size of the reading window an...
Preprint
Full-text available
Non-neural Machine Learning (ML) and Deep Learning (DL) are used to predict system failures in industrial maintenance. However, only a few studies have assessed the effect of varying the amount of past data used to make a prediction and the extension in the future of the forecast. This study evaluates the impact of the size of the reading window an...
Article
Full-text available
In Global North countries, persuasive apps supporting households in reducing their energy consumption are widespread, as promising policy tools for the energy and climate transition. Despite their growing diffusion, fed by the large-scale smart meter roll-out currently ongoing in many energy systems, rigorous studies providing evidence on their eff...
Article
Full-text available
Gravitational lensing is the relativistic effect generated by massive bodies, which bend the space-time surrounding them. It is a deeply investigated topic in astrophysics and allows validating theoretical relativistic results and studying faint astrophysical objects that would not be visible otherwise. In recent years, machine learning methods hav...
Chapter
Anomaly detection (AD) in numerical temporal data series is a prominent task in many domains, including the analysis of industrial equipment operation, the processing of IoT data streams, and the monitoring of appliance energy consumption. The life-cycle of an AD application with a Machine Learning (ML) approach requires data collection and prepara...
Article
Full-text available
Illegal landfills are sites where garbage is dumped violating waste management laws. Aerial images enable the use of photo interpretation for territory scanning and landfill detection but this practice is hindered by the manual nature of this task which also requires expert knowledge. Deep Learning methods can help capture the analysts’ expertise a...
Article
Full-text available
The diffusion of domotics solutions and of smart appliances and meters enables the monitoring of energy consumption at a very fine level and the development of forecasting and diagnostic applications. Anomaly detection (AD) in energy consumption data streams helps identify data points or intervals in which the behavior of an appliance deviates from...
Article
Full-text available
The application of Deep Neural Networks (DNNs) to a broad variety of tasks demands methods for coping with the complex and opaque nature of these architectures. When a gold standard is available, performance assessment treats the DNN as a black box and computes standard metrics based on the comparison of the predictions with the ground truth. A dee...
Article
Full-text available
The rapid development of remote sensing technologies and the availability of many satellite and aerial sensors have boosted the collection of large volumes of high-resolution images, promoting progress in a wide range of applications. As a consequence, Object detection (OD) in aerial images has gained much interest in the last few years. However, t...
Article
Full-text available
Object Detection requires many precise annotations, which are available for natural images but not for many non-natural data sets such as artworks data sets. A solution is using Weakly Supervised Object Detection (WSOD) techniques that learn accurate object localization from image-level labels. Studies have demonstrated that state-of-the-art end-to...
Chapter
The Distributed Ledger Technology (DLT) is one of the most durable results of virtual currencies, which goes beyond the financial sector and impacts business applications in general. Developers can empower their solutions with DLT capabilities to attain such benefits as decentralization, transparency, non-repudiability of actions and security and i...
Chapter
Anomaly detection is concerned with identifying rare events/observations that differ substantially from the majority of the data. It is considered an important task in the energy sector to enable the identification of non-standard device conditions. The use of anomaly detection techniques in small-scale residential and industrial settings can provi...
Chapter
Prescriptive maintenance has recently attracted a lot of scientific attention. It integrates the advantages of descriptive and predictive analytics to automate the process of detecting non nominal device functionality. Implementing such proactive measures in home or industrial settings may improve equipment dependability and minimize operational ex...
Preprint
Gravitational lensing is the relativistic effect generated by massive bodies, which bend the space-time surrounding them. It is a deeply investigated topic in astrophysics and allows validating theoretical relativistic results and studying faint astrophysical objects that would not be visible otherwise. In recent years Machine Learning methods have...
Chapter
Machine Learning (ML) tasks, especially Computer Vision (CV) ones, have greatly progressed after the introduction of Deep Neural Networks. Analyzing the performance of deep models is an open issue, addressed with techniques that inspect the response of inner network layers to given inputs. A complementary approach relies on ad-hoc metadata added to...
Preprint
The application of Deep Neural Networks (DNNs) to a broad variety of tasks demands methods for coping with the complex and opaque nature of these architectures. The analysis of performance can be pursued in two ways. On one side, model interpretation techniques aim at "opening the box" to assess the relationship between the input, the inner layers,...
Chapter
Full-text available
This chapter describes the various use cases of the SODALITE project in more detail. Each use case is representative of a unique infrastructure and operational environment supported by SODALITE: Snow (Cloud/OpenStack), Clinical Trials (HPC/Torque), and Vehicle IoT (Cloud + Edge / Kubernetes). Each section includes an overview of the specific challe...
Article
The concurrent development of applications requires reconciling conflicting code updates by different developers. Recent research on the nature of merge conflicts in open source projects shows that a significant fraction of merge conflicts have limited size (one or two lines of code) and are resolved with simple strategies that use code present in...
Article
Full-text available
Mapping the functional use of city areas (e.g., mapping clusters of hotels or of electronic shops) enables a variety of applications (e.g., innovative way-finding tools). To do that mapping, researchers have recently processed geo-referenced data with spatial clustering algorithms. These algorithms usually perform two consecutive steps: they cluste...
Article
Iconography in art is the discipline that studies the visual content of artworks to determine their motifs and themes and to characterize the way these are represented. It is a subject of active research for a variety of purposes, including the interpretation of meaning, the investigation of the origin and diffusion in time and space of representat...
Article
Full-text available
Iconography studies the visual content of artworks by considering the themes portrayed in them and their representation. Computer Vision has been used to identify iconographic subjects in paintings and Convolutional Neural Networks enabled the effective classification of characters in Christian art paintings. However, it still has to be demonstrate...
Article
Full-text available
Consumption-based feedback has been demonstrated to encourage water conservation behaviors. Smart meters and digital solutions can support customized feedback and reinforce behavioral change. Yet, most of the studies documenting water conservation effects induced by feedback and smart meter data visualization evaluate them in short-term experimenta...
Preprint
Iconography in art is the discipline that studies the visual content of artworks to determine their motifs and themes andto characterize the way these are represented. It is a subject of active research for a variety of purposes, including the interpretation of meaning, the investigation of the origin and diffusion in time and space of representati...
Chapter
Concurrent development requires the ability of reconciling conflicting updates to the code made independently. A specific case occurs when long living feature branches are integrated to a rapid changing code base. In this scenario, every integration test will require to manually resolve the same conflicts at every iteration. In this paper we propos...
Article
Full-text available
Virtual reality is a powerful interaction mechanism that holds the promise of engaging users, not only for entertainment, but also for social and environmental purposes. In this paper we present PeakLensVR, a virtual reality mobile application that enables users to capture panoramic mountain images with their mobile devices and later visualize such...
Chapter
Object detection and instance segmentation are major tasks in Computer Vision and have substantially progressed after the introduction of Deep Convolutional Neural Network (DCNN). Analyzing the performance of DCNNs is an open research issue, addressed with attention techniques that inspect the response of inner network layers to input stimuli. A co...
Article
FULL TEXT AVAILABLE! https://rdcu.be/b378l. ABSTRACT: Landform detection and analysis from Digital Elevation Models (DEM) of the Earth has been boosted by the availability of high-quality public data sets. Current landform identification methods apply heuristic algorithms based on predefined landform features, fine tuned with parameters that may...
Chapter
Energy efficiency requires a behavioral shift towards sustainable consumption. Such a change can be supported by persuasive IT applications, which employ a variety of stimuli to increase the energy literacy and awareness of consumers. We describe FunergyAR, an Augmented Reality digital game targeting children and their families. FunergyAR incorpora...
Article
Full-text available
Abstract Stimulating households to save energy with behaviour change support systems is a challenge and an opportunity to support efforts towards more sustainable energy consumption. The approaches developed so far, often either; do not consider the underlying behaviour change process in a systematic way, or do not provide a systematic linking of d...
Article
Full-text available
Abstract In this paper we present insights drawn from recent research projects aimed at developing visualization and gamification tools to stimulate individual behaviour change and promote energy and water saving. We address both the design of resource-saving programmes and the methodologies to assess their effectiveness. We conclude by presenting...
Article
Full-text available
Model Driven Development (MDD) requires proper tools to derive the implementation code from the application models. However, the integration of handwritten and generated code is a long-standing issue that affects the adoption of MDD in the industry. This article presents a model and code co-evolution approach that addresses such a problem a posteri...
Chapter
Artificial Intelligence on the edge is a matter of great importance towards the enhancement of smart devices that rely on operations with real-time constraints. We present PolimiDL, a framework for the acceleration of Deep Learning on mobile and embedded systems with limited resources and heterogeneous architectures. Experimental results show compe...
Conference Paper
Analyzing digital data to identify and classify landforms is an important task, which can contribute to improve the availability and quality of public open source cartography and to develop novel applications for tourism and environment monitoring. In the literature, several heuristic algorithms are documented for identifying the features of mounta...
Article
Full-text available
Computer and social sciences offer a wide range of tools to help face the world’s challenges arising in smart city scenarios and involving environment, energy, food, water, transportation, infrastructures, society, healthcare, education, governance, and economy. Indeed, purely technical solutions might be of little effect without proper considerati...
Conference Paper
Open Source Geographical Information Systems, such as OpenStreetMap (OSM), offer a valuable alternative to proprietary solutions for the development of voluntary environment monitoring systems. However, the quantity and quality of information stored in such systems must be carefully evaluated and the contributions of volunteers must be boosted by m...
Conference Paper
The use of IoT technologies (e.g. smart metering, smart home) and persuasive techniques to support energy-saving behaviour in households has been increasingly researched. Studies suggest that effective designs of such systems should incorporate different types of feedback with motivational techniques and energy saving advice [4]. But to be effectiv...
Article
Full-text available
In a world affected by the constant growth and concentration of the population in urban areas, the problem of preserving natural resources has become a priority. A promising approach to resource conservation is demand management, i.e., the ability to positively influence the behaviour of the population towards more sustainable consumption. Informat...
Conference Paper
Location-based mobile outdoor applications are powerful tools that can engage users in social and environmental tasks and support the emerging paradigm of citizen science. In this paper we present PeakLensVR, a virtual reality location-based mobile app that enables users to capture with their mobile phone panoramic mountain images and later visuali...
Article
Full-text available
Model Driven Engineering relies on the availability of software models and of development tools supporting the transition from models to code. The generation of code from models requires the unambiguous interpretation of the semantics of the modeling languages used to specify the application. This paper presents the formalization of the semantics o...
Article
Full-text available
Crowdsourcing marketplaces have emerged as an effective tool for high-speed, low-cost labeling of massive data sets. Since the labeling accuracy can greatly vary from worker to worker, we are faced with the problem of assigning labeling tasks to workers so as to maximize the accuracy associated with their answers. In this work, we study the problem...
Conference Paper
Outdoor augmented reality applications are an emerging class of software systems that demand the fast identification of natural objects, such as plant species or mountain peaks, in low power mobile devices. Convolutional Neural Networks (CNN) have exhibited superior performance in a variety of computer vision tasks, but their training is a labor in...
Conference Paper
Full-text available
Rapid evolution and flexibility are the key of modern web application development. Rapid Prototyping approaches try to facilitate evolution by reducing the time between the elicitation of a new requirement and the evaluation of a prototype by both developers and customers. Software generation, with disciplines such as Software Product Lines Enginee...
Article
SmartH2O is a software platform that creates a virtuous feedback cycle between water users and the utilities, providing users information on their consumption in quasi real time, and thus enabling water utilities to plan and implement strategies to reduce/reallocate water consumption. The SmartH2O platform adopts a gamification paradigm to motivate...
Conference Paper
This paper presents the research objectives of the enCOMPASS project, which aims at implementing and validating an integrated socio-technical approach to behavioural change for energy saving. To this end, innovative user-friendly digital tools will be developed to 1) make energy data consumption available and understandable for different types of u...
Conference Paper
Software product lines allow users with little development experience to configure and generate applications. On the web this approach is becoming more and more popular due to the low time required to bring a new release to the final users. The architecture of web applications though require complex development environments in order to allow users...
Article
Full-text available
Snow is a key component of the hydrologic cycle in many regions of the world. Despite recent advances in environmental monitoring that are making a wide range of data available, continuous snow monitoring systems that can collect data at high spatial and temporal resolution are not well established yet, especially in inaccessible high-latitude or m...
Conference Paper
This paper merges multimedia and environmental research to verify the utility of public web images for improving water management in periods of water scarcity, an increasingly critical event due to climate change. A multimedia processing pipeline fetches mountain images from multiple sources and extracts virtual snow indexes correlated to the amoun...
Conference Paper
Full-text available
The advent of connected mobile devices has caused an unprecedented availability of geo-referenced user-generated content, which can be exploited for environment monitoring. In particular, Augmented Reality (AR) mobile applications can be designed to enable citizens collect observations, by overlaying relevant meta-data on their current view. This c...
Preprint
Full-text available
Snow is a key component of the hydrologic cycle in many regions of the world. Despite recent advances in environmental monitoring are making a wide range of data available, continuous snow monitoring systems able to collect data at high spatial and temporal resolution are not well established yet, especially in inaccessible high latitude or mountai...
Conference Paper
Full-text available
Stimulating users to save water is a challenge and an opportunity for water demand management. Existing ICT-based systems for behavioural change often do not consider the underlying behavioural determinants in a systematic way. This paper discusses the design of the behavioural change and incentive model for the SmartH2O system, combining smart met...
Conference Paper
Methodological frameworks guide the design of digital learning game based on well founded learning theories and instructional strategies. This study presents a comparison of five methodological frameworks for digital learning game design, highlighting their similarities and differences. The objective is to support the choice of an adequate framewor...
Conference Paper
Outdoor augmented reality applications project information of interest onto views of the world in real-time. Their core challenge is recognizing the meaningful objects present in the current view and retrieving and overlaying pertinent information onto such objects. In this paper we report on the development of a framework for mobile outdoor augmen...
Conference Paper
The demo presents SnowWatch, a citizen science system that supports the acquisition and processing of mountain images for the purpose of extracting snow information, predicting the amount of water available in the dry season, and supporting a multi-objective lake regulation problem. We discuss how the proposed architecture has been rapidly prototyp...
Chapter
This chapter illustrates the IFML constructs for representing the general organization of the interface independently of the content published in the view. Two types of organizations are possible: one typical of web applications, where multiple peer-level ViewContainers embody the content and navigation of the interface; one typical of desktop, mob...
Chapter
The chapter addresses the specification of the content and navigation aspects of the interface and shows how to use ViewContainers, Events, NavigationFlows, and DataFlows to describe many configurations. The readability of models is enhanced by using more specific ViewComponents, such as List and Details, which make diagram more understandable and...
Chapter
The chapter discusses the IFML concept of Action, which describes a black-box component that embodies arbitrary business logic triggered from the interface. Actions can be connected to interface elements with navigation and data flows to enable parameter passing. The chapter illustrates several design patterns involving actions, mostly for updating...
Chapter
This chapter illustrates the role of the extension mechanism natively provided by the Interaction Flow Modeling Language (IFML). The basic constructs of the language can be extended, to better adhere to the terminology and concepts of a specific class of applications and to improve model checking and code generation. The authors show the extension...
Chapter
This chapter discusses some of the aspects of the IFML language design: the formal definition of the concepts in the IFML metamodel, the model exchange format, the executability, and the integration with other models representing other aspects or perspectives upon the system besides user interaction.
Chapter
This chapter exemplifies the support to IFML model-driven development with the help of a specific tool, called WebRatio. WebRatio is a composite application development tool that covers not only the front-end design, but also domain modeling, business logic modeling, and process modeling, thus providing an end-to-end approach to model-driven develo...
Chapter
This chapter aims at addressing the typical problems of user interface (UI) design, by providing a reasoned categorization of classical user interaction patterns in modern interfaces. Each pattern is described by its IFML model, an exemplary UI rendering and a textual explanation of its behavior.
Chapter
This chapter delves into the discussion on how to map the platform-independent IFML models into specific technological platforms. Ideally, the mapping to the implementation layer could be illustrated for any software architecture that supports user’s interactivity. For space reasons, this chapter restricts the illustration to four main categories o...
Chapter
This chapter addresses domain modeling. IFML does not prescribe a specific domain modelling language but can be interfaced to any notation preferred that expresses the objects and associations of the application domain. The chapter employs UML class diagrams, and briefly recaps their main features for structural modeling. It discusses design patter...
Chapter
This chapter provides a sample of realistic applications, inspired from real-world popular ones, and describes how they can be modeled in using the Interaction Flow Modeling Language (IFML). More precisely, the chapter covers two large modeling examples: the first is a mobile app tailored to smartphones, providing an online photo and video-sharing...
Chapter
This chapter provides a bird's eye view of IFML. The chapters presents the main language concepts: ViewContainers, ViewComponents, Events, InteractionFlows, Parameters, ParameterBindings and Actions. IFML concepts are referred to the elements of the Model-View-Controller design pattern. These concepts are illustrated in a small, yet complete, examp...
Conference Paper
We present a study that reveals a significant statistical bias in the distributions of geolocated and non-geolocated social data. We state that this bias affects the real performance of social geolocation algorithms and can impair the results of these algorithms, which are commonly trained and tested on datasets consisting of crawled geolocated dat...
Article
Full-text available
Despite the success of crowdsourcing marketplaces, fully harnessing their massive workforce remains challenging. In this work we study the effect on crowdsourc-ing campaigns of different feedback and payment strategies. Our results reveal the joint effect of feedback and payment on the quality and quantity of the outcome.
Conference Paper
Full-text available
Despite the success of crowdsourcing marketplaces, fully harnessing their massive workforce remains challenging. In this work we study the effect on crowdsourc-ing campaigns of different feedback and payment strategies. Our results reveal the joint effect of feedback and payment on the quality and quantity of the outcome.
Conference Paper
Full-text available
This paper considers the structural similarities in approaches and lessons learned in the development of applications for behavior change in water and energy saving. We show how the domains of water and energy are related and propose a first set of design guidelines for building such solutions, especially regarding visualization and gamification of...
Conference Paper
Full-text available
The increasing popularity of social media articles and micro-blogging systems is changing the way online information is produced: users are both content publishers and content consumers. Since information is produced and shared by common users, who usually have a limited domain knowledge, and due to an exponential growth of the available informatio...
Article
In this paper we study the problem of estimating snow cover in mountainous regions, that is, the spatial extent of the earth surface covered by snow. We argue that publicly available visual content, in the form of user generated photographs and image feeds from outdoor webcams, can both be leveraged as additional measurement sources, complementing...