Machine learning methods are increasingly being used for the analysis of Earth observation data collected through remote-sensing, for instance in tasks such as land cover mapping. However, problems that involve combining remotely-sensed data with volunteered geographical information (e.g., ground-level photos or data from sources such as the OpenStreetMap) are only now starting to be explored, and they still involve a number of practical challenges. Over the recent years, image classification, segmentation or super-resolution, leveraging deep neural networks, have also become increasingly popular. These methods have been reported to result in impressive performance gains, when applied to problems related to processing natural images. Deep learning methods can also have several applications within GIScience research, motivating the design of tailored methods.
Within the MIMU project (i.e., acronym for MIning MUlti-source and MUlti-modal geo-referenced information), the researchers to be hired will work on the use of deep learning approaches for the discovery and mapping of innovative geographic knowledge, through the analysis and processing of large-scale volunteered data (e.g., geo-referenced multimedia contents) in combination with more traditional sources (e.g., remote-sensing products available in the context of initiatives like ESA's Sentinel/Copernicus programme). The complex relations between the different types of information, as well as the temporal and geographical dimensions of the data, introduce new challenges that will explored throughout the project, in an attempt to go beyond the current state-of-the-art. Research within the project will contribute to the development of spatially explicit deep learning methods, envisioning a variety of practical applications (e.g., land cover mapping, remote sensing image captioning and visual question answering, forecasting with remote sensing data, etc.).
One of the researchers to be hired in the project will (research position A) will develop work with a focus on image processing (e.g., considering tasks such as land cover mapping, semantic segmentation of high resolution of remote sensing imagery, or forecasting with remote sensing data), while the other (research position B) will develop work focused on jointly processing textual and visual contents (e.g., considering tasks such as remote sensing image retrieval, captioning and visual question answering, geocoding images and textual contents, etc.).

Dissemination of results, project management, scientific advising and preparation of scientific meetings.
Coordination of the project’s use cases and student advising.
Design and implementation of location proof techniques.
Definition of data schemas and application programming interfaces.
Development of witness location models; integration with identity services.

Development of predictive modules for the number and type of medical emergencies in space and time using advanced techniques for pattern discovery in data. Finding correlations of medical emergencies with external data such as events or weather information.

The researcher will have to work with healthcare professionals and do the following tasks:
•	Apply user-centred design methodologies to collect requirements for rehabilitation technologies in clinical settings;
•	Design and develop interactive applications for stroke patients that leverage data from inertial sensors and 3D cameras (e.g. Kinect);
•	Develop tools that support physiotherapists' activities in clinical rehabilitation settings;
•	Prepare and conduct field studies to assess the effectiveness of developed technologies;
•	Supervise students at MSc and PhD level;
•	Submit, at least, three papers to top-level venues.

This research work aims to develop a fault-tolerant embedded system within the scope of project SARRROCA. The research work will involve designing novel hardware architectures, and study and implement novel fault-tolerant mechanisms. Expected outputs of this work are the implementation of prototypes for real-time fault-tolerant architectures.