Learning from Demonstrations (LfD) via Behavior Cloning (BC) works well on multiple complex tasks. However, a limitation of the typical LfD approach is that it requires expert demonstrations for all scenarios, including those in which the algorithm is already well-trained. The recently proposed Learning from Interventions (LfI) overcomes this limitation by using an expert overseer. The expert overseer only intervenes when it suspects that an unsafe action is about to be taken. Although LfI significantly improves over LfD, the state-of-the-art LfI fails to account for delay caused by the expert's reaction time and only learns short-term behavior. We address these limitations by 1) interpolating the expert's interventions back in time, and 2) by splitting the policy into two hierarchical levels, one that generates sub-goals for the future and another that generates actions to reach those desired sub-goals. This sub-goal prediction forces the algorithm to learn long-term behavior while also being robust to the expert's reaction time. Our experiments show that LfI using sub-goals in a hierarchical policy framework trains faster and achieves better asymptotic performance than typical LfD.
Energy Autonomous Wearable Sensors (EAWS) have attracted a large interest due to their potential to provide reliable measurements and continuous bioelectric signals, which help to reduce health risk factors early on, ongoing assessment for disease prevention, and maintaining optimum, lifelong health quality. This review paper presents recent developments and state-of-the-art research related to three critical elements that enable an EAWS. The first element is wearable sensors, which monitor human body physiological signals and activities. Emphasis is given on explaining different types of transduction mechanisms presented, and emerging materials and fabrication techniques. The second element is the flexible and wearable energy storage device to drive low-power electronics and the software needed for automatic detection of unstable physiological parameters. The third is the flexible and stretchable energy harvesting module to recharge batteries for continuous operation of wearable sensors. We conclude by discussing some of the technical challenges in realizing energy-autonomous wearable sensing technologies and possible solutions for overcoming them.
We tackle the issue of measuring and analyzing the visitors' dynamics in crowded museums. We propose an IoT-based system -- supported by artificial intelligence models -- to reconstruct the visitors' trajectories throughout the museum spaces. Thanks to this tool, we are able to gather wide ensembles of visitors' trajectories, allowing useful insights for the facility management and the preservation of the art pieces. Our contribution comes with one successful use case: the Galleria Borghese in Rome, Italy.
Many decision-making processes have begun to incorporate an AI element, including prison sentence recommendations, college admissions, hiring, and mortgage approval. In all of these cases, AI models are being trained to help human decision makers reach accurate and fair judgments, but little is known about what factors influence the extent to which people consider an AI-infused decision-making process to be trustworthy. We aim to understand how different factors about a decision-making process, and an AI model that supports that process, influences peoples' perceptions of the trustworthiness of that process. We report on our evaluation of how seven different factors -- decision stakes, decision authority, model trainer, model interpretability, social transparency, and model confidence -- influence ratings of trust in a scenario-based study.
People involved in mass emergencies increasingly publish information-rich contents in online social networks (OSNs), thus acting as a distributed and resilient network of human sensors. In this work, we present HERMES, a system designed to enrich the information spontaneously disclosed by OSN users in the aftermath of disasters. HERMES leverages a mixed data collection strategy, called hybrid crowdsensing, and state-of-the-art AI techniques. Evaluated in real-world emergencies, HERMES proved to increase: (i) the amount of the available damage information; (ii) the density (up to 7x) and the variety (up to 18x) of the retrieved geographic information; (iii) the geographic coverage (up to 30%) and granularity.
The interactions and activities of hundreds of millions of people worldwide are recorded as digital traces every single day. When pulled together, these data offer increasingly comprehensive pictures of both individuals and groups interacting on different platforms, but they also allow inferences about broader target populations beyond those platforms, representing an enormous potential for the Social Sciences. Notwithstanding the many advantages of digital traces, recent studies have begun to discuss the errors that can occur when digital traces are used to learn about humans and social phenomena. Incidentally, many similar errors also affect survey estimates, which survey designers have been addressing using error conceptualization frameworks such as the Total Survey Error Framework. In this work, and leveraging the systematic approach of the Total Survey Error Framework, we propose a conceptual framework to diagnose, understand and avoid errors that may occur in studies that are based on digital traces of humans.