Project Specific Grants / Reserved topic scholarships 2018 - 2nd call
Centro Ricerche Fiat S.c.p.a., Fondazione Bruno Kessler (FBK) and EIT Digital
B2 - Vehicle to Everything: opportunities and constraints in leading the automotive world to 5G (1 grant)
Intelligent Transport Systems (ITS) applications are supported by several communication technologies, each one with its frequency range and specific features. Evaluating the performance of different network options for V2X communication that ensure optimal utilization of resources is a prerequisite when designing and developing robust wireless networks for ITS applications. 5G networks are expected to leverage on virtualization of network resources in order to serve over the same infrastructures applications and services characterized by highly heterogeneous requirements, the so called verticals. The thesis will investigate the potentialities introduced by the 5G network for the automotive domain, identifying use cases and scenarios, and deriving requirements for the M(V)NO. The identified solution will be experimentally validated in a lab environment (Hardware in the loop or simulated scenarios) and in a more realistic conditions.
The scholarship is subject to the acceptance of nondisclosure agreements and the assignment of the outcomes.
The beneficiary of the scholarship will be included in the EIT Digital I&E Education course, where he/she will receive training on aspects of entrepreneurship and innovation through participation in four courses organised by the Doctoral Training Center of EIT Digital. He/she will also be required to carry out a 6-month study or research period abroad, within the research theme. In addition, the grant beneficiary must obtain a Business Development Experience at Centro Ricerche Fiat S.c.p.a. where scientific work will be integrated into entrepreneurship and innovation activities.
The beneficiary of the scholarship will carry out the research activity at the EIT Co-location Center in Trento, using its laboratories and equipment, without prejudice to the commitments related to educational activity of the DISI PhD programme (as distinct from the work experience and the research side) and excluding the training period abroad according to article 9 of the Ministerial Decree 45/2013, as well as the period dedicated to the Business Development Experience.
Contact: riggio [at] fbk.eu
DISI - Department of Information Engineering and Computer Science
C1 - Quantum Annealing for SAT Solving(1 grant)
This PhD research is sponsored by and in collaboration with D-Wave Systems Inc. (http://www.dwavesys.com). The goal is to investigate the usage of D-Wave's quantum processors (QPUs) to solve hard SAT and related NP-Hard problems by exploiting quantum effects. The ultimate goal is to solve problems which are currently out of the reach of state-of-the-art SAT solvers (e.g., from cryptanalysis). The proposed research is to develop effective and efficient encoding procedures from SAT to problems which fit into, and can be solved by, D-Wave's QPUs. These procedures will be presumably based on Satisfiability Modulo theories, atuomated-reasoning and graph-manipulation techniques.
The scholarships is subject to the acceptance of the assignment of the ownership of intellectual property of the research results.
A more detailed description is available as http://disi.unitn.it/rseba/DWAVE-Recruit-phd.txt
Please contact proposer for further explaination
D1 - Development of Machine Learning and Pattern Recognition Methods for the Analysis of Big Data from Space (1 grant)
The activity that will be developed is related to the definition, the design, the development and the validation of machine learning and pattern recognition techniques for the analysis of big data from space. The activities will address different topics related to the main methodological problems to be solved for the analysis of the huge archives of remote sensing data that are acquired by Earth Observation satellites. The main topics include: 1) deep learning architectures for the automatic classification of satellite images; 2) data analysis techniques for the automatic extraction of semantic information from data (e.g. content based retrieval); 3) data fusion techniques for the integration of multisensor and multisource data. The focus on one or more of the above-mentioned topics will be agreed with the selected candidate. The PhD student will work at the Remote Sensing Laboratory of the University of Trento.
For more information on the activity of RSLab refer to: http://rslab.disi.unitn.it/ or contact lorenzo.bruzzone [at] unitn.it
FBK - Fondazione Bruno Kessler
(*) All scholarships funded by FBK are subject to the acceptance of the assignment of the ownership of intellectual property of the research results.
A1 - Automatic Estimation of Spoken Second Language Proficiency (1 grant) (*)
The problem of automatic scoring of second language (L2) learning proficiency has been largely investigated in the past in the framework of computer assisted language learning. Approaches have been proposed for the two input modalities to scoring models, i.e. written and spoken, and in both cases they rely on features representing specific competences of the human learners, e.g. grammar correctness, comprehension, fluency, quality of pronunciation etc. The final goal is to measure language proficiency according to some standard scale; a well known L2 scale is the Common European Framework of Reference for Languages, that defines 6 levels of proficiency (A1, A2, B1, B2, C1, C2) and is widespread also out of Europe. The objective of the thesis will be to investigate approaches for the automated assessment of spoken language proficiency using machine learning technologies. These will make use of automatic speech recognition (ASR) of spoken input to feed several types of deep neural networks. The addressed task is very challenging and poses a certain number of problems to solve. From the ASR perspective major difficulties are represented by both the recognition of non-native speech and the recognition of spontaneous speech. Concerning proficiency estimation the major work will address the design of suitable features, from both the automatic transcription and the voice signal, to score different types of proficiency indicators.
Contact: matasso [at] fbk.eu - falavi [at] fbk.eu
A2 - Computation and network neuroscience (1 grant) (*)
Building on empirical evidence for the strict relationship between human brain and cognition, generally referred to as mind, this project between FBK and APSS (Azienda Provinciale per i Servizi Sanitari della Provincia Autonoma di Trento) aims at integrating neuroimaging and neuropsychological analyses to quantify how mental pathways are altered in diseases and how they relate to structural and functional neural pathways.
Contact: merler [at] fbk.eu - mdedomenico [at] fbk.eu
A3 - Middleware and technologies for real time analysis of IoT streams with machine learning algorithms (1 grant) (*)
The Internet of Things (IoT) is an emerging paradigm that sees millions of sensors and actuators involved in the monitoring and management of systems combining physical machines, devices and humans. The capability to deliver embedded and distributed intelligence allows to provide smarter IoT objects capable to autonomously recognize patterns and environmental conditions and of properly react without the need for centralized control or for human intervention, improving responsiveness and effectiveness of delivered IoT systems.
The objective of the PhD is the study, analysis and experimentation on middleware and technologies that support the execution of machine and deep learning algorithms on IoT streams providing both short term, closed-loop decision making capabilities at device level, and scalable long term analysis capabilities of a distributed IoT system, leveraging on the different computational capabilities distributed in the various portions of the whole IoT infrastructure, both in the cloud and in the edge of the network (in the IoT gateways and in the IoT devices themselves).
Contact: mvecchio [at] fbk.eu
A4 - Modeling and Predicting Social and Financial Wellbeing (1 grant) (*)
The ability of understanding and predicting the social and financial wellbeing for individuals, companies, and societies is of interest to economists, policy designers, financial institutions, and the individuals themselves. Humans have often been described as socio-economic beings given that their financial and economic behavior and, more in general their well-being, is intricately connected with their social behavior. In this project, the goal is merging approaches from network science and machine learning and using data on social interactions, mobility routines and purchase behaviors in order to develop methods for quantify financial and social wellbeing. The Ph.D. project will be conducted within the FBK MobS research unit but with collaborations with the FBK CoMuNe research unit and the Human Dynamics group at MIT Media Lab.
Contact: lepri [at] fbk.eu
A5 - End-to-end Dialogue Models for Conversational Agents and ChatBots(1 grant) (*)
Conversational agents are designed to interact with users in multiple domains on several topics using natural language. Many chatbots have been deployed on the Internet (social media, e-commerce websites, just to mention a few) for the purpose of seeking information, question answering, coaching tasks, online shopping, and so on. Usually these applications work in a strictly limited domain with a clear and well defined dialogue
structure, with little adaptation capabilities to the contextual and social situation. The goal of this PhD Thesis is to overcome the shortcoming of traditional goal-oriented applications that require a lot of domain-specific handcrafting, which hinders scaling up to new domains and languages. To do so, end-to-end approaches, in which all components are trained from the dialogs themselves, can be used to incorporate several dialogue and domain features.
Contact: magnini [at] fbk.eu
B1 - Hardware and software solutions for blockchain-based IoT applications (1 grant) (*)
Blockchain can offer IoT devices a playground where they can be identified without the need of involving central trusted authorities (decentralised identity control) and the possibility to operate and interact within a trust-less environment. One of the big challenges of the use of blockchain technologies in combination with IoT is how to provide solutions capable to guarantee at hardware and software level a native support for blockchain and distributed ledger technologies to provide IoT devices with the capability to perform autonomous transactions that result to be intrinsically secure and scalable.
The objective of the PhD is the study, analysis and experimentation of how the combination of hardware, cryptography and software solutions can provide foundations for the creation of infrastructures enabling IoT devices to perform autonomous secured transactions in a trustless environment.
Contact: mvecchio [at] fbk.eu
C2 - Gamification for Smart Cities and Communities (1 grant) (*)
In recent years, gamification has been successfully applied to increase people’s engagement, and to leverage contemporary ICT to promote participation and positive behavioral changes. FBK develops approaches and technologies for the gamification of dynamic and open-bounded socio-technical systems, such as Smart Communities and Smart Cities. Our research covers the whole game lifecycle, from game design to development, deployment, execution, monitoring and analytics. This topic offers several open research challenges: i) definition of concepts, models, and languages for the specification and design support of complex games that can apply to, and can be reused in, a variety of Smart City domains (e.g., sustainable mobility, energy conservation, participatory governance) and settings (e.g., various cities and communities); ii) AI techniques (e.g., machine learning, reccommender systems, planning) supporting the procedural generation of game content that adapts the game experience of players to meet dynamic game objectives, matches the personal player profile and ensures the engagement and retainment of players.
Contact: marconi [at] fbk.eu
C3 - Advanced formal methods for the development and verification of an autonomous vehicle in the railway domain (1 grant) (*)
As part of an industrial project, FBK and Rete Ferroviaria Italiana (RFI) have a joint collaboration to formally develop an autonomous railway vehicle, used for maintenance of the railway line. The vehicle must be able to inspect line sections in an autonomous manner or guided from the ground, according to a pre-established journeyplan. In particular, this study will focus on the ATO (Autonomous Train Operation) sub-system. The object of the present study is the development and adoption of extended formal design methodologies for ATO. These methodologies will be applied to the design of the (ground and on-board) control logic of ATO, and to the realization of a demonstrator for the remote guidance system. Formal design methodologies will cover the analysis of the functional and non-functional requirements of the system, the modeling, verification and validation of its control logic, and finally the implementation of the software in C code, in accordance with the standard EN50128: 2011. Qualifying aspects of the activity are the modeling, to be made using formalisms such as architectural diagrams and finite state machines, and the validation and formal verification of the system and of its implementation. The study will integrate traditional software engineering methodologies based on testing, and automated verification and validation techniques based on model checking. The study will also take into consideration safety aspects and non-functional aspects such as energy management.
Contact: cimatti [at] fbk.eu
D2 - IoT for Smart Cities and Communities (1 grant) (*)
The Internet of Things, including smart objects, wearables and wireless sensor networks, is becoming a key technology to enable Smart Cities and Communities. Key scenarios require connecting people and their environments through smart services supported by intelligent devices that observe and provide enriched contextual information. As such devices increasingly pervade physical spaces, challenges of sustainability arise in terms of scalability, device lifetime, and management of the massive amounts of generated data. We approach these challenges starting at the "smart thing" level, exploiting local processing to limit communication and adapting low power wireless communication protocols such as BLE and LoRa to offload information at low cost.
Motivated by the challenges of Smart Cities and Communities, this research aims to address a range of challenges, to be balanced based on the curriculum of the candidate:
- define the "next smart thing" from an energy efficient embedded systems perspective, considering hardware architecture at system level and considering requirements arising from Smart Communities;
- explore innovative machine learning approaches to be embedded on resource-constrained devices;
- face the energy efficiency challenge of IoT devices from the perspective of wireless communication, considering standard and non-standard communication stacks. Preferential background in both signal processing or machine learning, sensor technologies and microcontroller/embedded programming if the candidate is willing to tackle the described research from the point of view of smart sensors/smart computing at the edge. Alternatively experience with wireless technologies, network simulation environment and tools if the candidate is more interested on the wireless communication side.
Contact: murphy [at] fbk.eu - efarella [at] fbk.eu
D3 - Fusion of remote sensing and citizen science information for geospatial products (1 grant) (*)
An actual scientific challenge is the design of a framework to create a strong connection between humans and machines in order to improve higher quality of geospatial products. In recent years a fast increment of activities has been observed on both: 1) citizen science and its expansion by the reuse of crowdsourcing data from other projects (e.g. OpenStreetMap). This happened in many scenarios including Earth Observation; and 2) information extraction from remotely sensed data acquired by satellite-and airborne-based platforms. Remote sensing state of the art is wide from both methodological and application viewpoints. Both research fields are
acquiring more and more relevance in our society.
In this context the Ph.D. candidate will work on the design of novel methods for the fusion of remote sensing technology and citizen science for: i) novel products and services that may gain from the joint exploitation of the two complementary data sources; and ii) cross-validate remote sensing products by citizen science data or viceversa.
Candidates are preferred having a background in geoinformation analysis, image processing, computer science, physics and simila.
Contact: bovolo [at] fbk.eu - napo [at] fbk.eu
D4 - Multi-sensor data fusion for terrestrial and indoor 3D modeling (1 grant) (*)
The complexity of man-made indoor scenarios or architectures deserves powerful, reliable, replicable, accurate and effective 3D surveying methodology to correctly model all small geometric and radiometric features. Current approaches rely on a single techniques or sensor or perform an integration of the derived 3D data at the end of the process. The aim of the research is to investigate how to efficiently integrate / fuse various acquisition sensors (cameras, laser scanners, GNSS, etc.) and data (images, point clouds, etc.) at raw data level for modelling terrestrial and indoor environments. The fusion needs to consider the available data and sensors, exploiting intrinsic properties and synchronizing methods. 3D modeling should rely also on new machine learning methods for 2D and 3D data to speed up and automatize the processing.
Contact: remondino [at] fbk.eu
IIT - Fondazione Istituto Italiano di Tecnologia
A6 - Domain adaptive visual detection (1 grant)
The ability to detect objects, persons and more in general visual entities in images and videos is a key ability for seeing machines. A very desirable feature is robustness, namely the ability of detection algorithms to work at test time on data perceptually very different from those used during training in terms of illumination conditions, scale, viewpoint, context and so forth. Domain adaptation and generalization methods have historically attempted to address these problems mostly within the object classification framework. The goal of this PhD project is to frame the domain adaptation problem within the visual detection framework and develop algorithms computationally efficient, able to work robustly for detection of various visual patterns, in indoor and outdoor scenarios, by leveraging over domain adaptation and generalization techniques.
Contact: Barbara.Caputo [at] iit.it - e.ricci [at] unitn.it
OSRAM GmbH, Department of Information Engineering and Computer Science (DISI) and EIT Digital
A7 - Domain adaptation for people detection, re-identification and pose estimation (1 grant) [additional reserved topic scholarship]
Computer vision and machine learning solutions for people detection and re-identifications have been proposed in the past but the main limitation is that they depend on specific data which the models were trained on. To address this, project addresses research on domain adaptation in order to investigate the models, learning and data strategies which enable generalization. We will consider algorithms for people detection, re-identification and pose estimation, which we will generalize to novel environments, e.g. different people and illumination, different cameras and optics, and different setups. This is particularly relevant to the EIT Digital city action line since it involves analytics for decision making, as understanding the people’s locations/pose underlies the recognition of actions and the social activities. The research is also relevant to resilient-safety city, as it may support the identification/tracking of wanted individuals or accessing limited-access areas. We will consider the successful innovation transfer to a number of industrial applications directly relevant to OSRAM, including, but not limited to, office and retail.
The scholarship is subject to the acceptance of nondisclosure agreements and the assignment of the outcomes.
The beneficiary of the scholarship will be included in the EIT Digital I&E Education course, where he/she will receive training on aspects of entrepreneurship and innovation through participation in four courses organised by the Doctoral Training Center of EIT Digital. He/she will also be required to carry out a 6-month study or research period abroad, within the research theme. In addition, the grant beneficiary must obtain a Business Development Experience at OSRAM where scientific work will be integrated into entrepreneurship and innovation activities.
The beneficiary of the scholarship will carry out the research activity at the EIT Co-location Center in Trento, using its laboratories and equipment, without prejudice to the commitments related to educational activity of the DISI PhD programme (as distinct from the work experience and the research side) and excluding the training period abroad according to article 9 of the Ministerial Decree 45/2013, as well as the period dedicated to the Business Development Experience.
Contact: niculae.sebe [at] unitn.it - fabio.galasso [at] gmail.com