Ricerc@Sapienza

Il lavoro si propone di individuare ed esaminare i problemi giuridici posti dai processi decisionali automatizzati, come regolati dal GDPR. In particolare, il lavoro si concentra sul tema dell'efficacia e dell'effettività degli strumenti di tutela attributi dal GDPR all'individuo sottoposto al processo decisionale automatizzato, evidenziando i limiti del principio della trasparenza - tradizionalmente inteso - rispetto ai sistemi dell'intelligenza artificiale e ai modelli di machine learning.

Starting from 2013, the number of unvaccinated people alarmingly increased in Italy; therefore, in 2017 a new Vaccine National Plan was approved. Healthcare workers (HCWs), especially public health professionals (PHPs, i.e., workers in in the sector of hygiene and preventive medicine), have an important role in informing and promoting vaccinations.

In the last years, the design and implementation of web-based education systems has grown exponentially, spurred by the fact that neither students nor teachers are bound to a specific location and that this form of computer-based education is virtually independent of any specific hardware platform. These systems accumulate a large amount of data: educational data mining and learning analytics are the two much related fields of research with the aim of using these educational data to improve the learning process.

We present a novel approach for LDA (Latent Dirichlet Allocation) topic reconstruction. The main technical idea is to show that the distribution over the documents generated by LDA can be transformed into a distribution for a much simpler generative model in which documents are generated from the same set of topics but have a much simpler structure: documents are single topic and topics are chosen uniformly at random.

In this paper we show the importance of the head pose estimation in the task of trajectory forecasting. This cue, when produced by an oracle and injected in a novel socially-based energy minimization approach, allows to get state-of-the-art performances on four different forecasting benchmarks, without relying on additional information such as expected destination and desired speed, which are supposed to be know beforehand for most of the current forecasting techniques.

Recent approaches on trajectory forecasting use tracklets to predict the future positions of pedestrians exploiting Long Short Term Memory (LSTM) architectures. This paper shows that adding vislets, that is, short sequences of head pose estimations, allows to increase significantly the trajectory forecasting performance. We then propose to use vislets in a novel framework called MX-LSTM, capturing the interplay between tracklets and vislets thanks to a joint unconstrained optimization of full covariance matrices during the LSTM backpropagation.

Person search has recently gained attention as the novel task of finding a person, provided as a cropped sample, from a gallery of non-cropped images, whereby several other people are also visible. We believe that i. person detection and re-identification should be pursued in a joint optimization framework and that ii. the person search should leverage the query image extensively (e.g. emphasizing unique query patterns). However, so far, no prior art realizes this. We introduce a novel query-guided end-to-end person search network (QEEPS) to address both aspects.

In this work, we explore the correlation between people trajectories and their head orientations. We argue that people trajectory and head pose forecasting can be modelled as a joint problem. Recent approaches on trajectory forecasting leverage short-term trajectories (aka tracklets) of pedestrians to predict their future paths. In addition, sociological cues, such as expected destination or pedestrian interaction, are often combined with tracklets.

Neural network compression has recently received much attention due to the computational requirements of modern deep models. In this work, our objective is to transfer knowledge from a deep and accurate model to a smaller one.

Machine learning algorithms are responsible for a significant amount of computations. These computations are increasing with the advancements in different machine learning fields. For example, fields such as deep learning require algorithms to run during weeks consuming vast amounts of energy. While there is a trend in optimizing machine learning algorithms for performance and energy consumption, still there is little knowledge on how to estimate an algorithm’s energy consumption.