Perceiving the 3d world from single images

Depth represents a crucial piece of information in many practical applications, such as obstacle avoidance and environment mapping. This information can be provided either by active sensors, such as LiDARs, or by passive devices like cameras. A popular passive device is the binocular rig, which allows triangulating the depth of the scene through two synchronized and aligned cameras. However, many devices that are already available in several infrastructures are monocular passive sensors, such as most of the surveillance cameras. The intrinsic ambiguity of the problem makes monocular depth estimation a challenging task. Nevertheless, the recent progress of deep learning strategies is paving the way towards a new class of algorithms able to handle this complexity. This work addresses many relevant topics related to the monocular depth estimation problem. It presents networks capable of predicting accurate depth values even on embedded devices and without the need of expensive ground-truth labels at training time. Moreover, it introduces strategies to estimate the uncertainty of these models, and it shows that monocular networks can easily generate training labels for different tasks at scale. Finally, it evaluates off-the-shelf monocular depth predictors for the relevant use case of social distance monitoring, and shows how this technology allows to overcome already existing strategies limitations.

Beyond weaponry and socialist internationalism: Soviet engagement with postcolonial state-building in the Arab Middle East, 1954-1966

This research investigates the Soviet Union’s role in guiding state-building processes of postcolonial Arab countries of the Middle East, leading them to adopt economic and political elements of the socialist-Leninist models of development. The widespread narrative depicts the Soviets as having failed to export communism in those states and, therefore, as having failed to bring them closer to Moscow’s sphere of influence and values. However, various Soviet archives suggest a different reality. As the Cold War burst forth, between the mid-1950s and mid-1960s, contacts between Soviet and Arab officials were not just incredibly frequent but they went to the core of all main issues of socio-economic development of these transforming countries: party politics, institution building, agrarian reforms, industrialisation, security sector reforms, etc. The research focuses on a period that may be labelled as ‘the launching phase’ of the Soviet Middle East policy, which established a long-lasting framework for the Soviet-Arab dialogue. It also places significant attention on the ‘personal dimension’ of such a dialogue, showing how Moscow’s influence went hand in hand with the ability of Soviet leaders and diplomats to establish relations of personal trust with postcolonial Arab élites. A selected number of Arab countries are examined: Egypt, Iraq and Syria. For each of these countries, a limited period of time will be taken into consideration, when Soviet influence reached its peak and state-building policies might have drawn from the Soviet model (for Egypt 1954-1958; for Iraq 1958-1963; for Syria 1961-1966). On the one hand, the analysis of specific case-studies will allow to investigate the relationship between Moscow and each of these new Arab regimes; on the other, such a large geographical scope will permit to grasp the elements and the objectives of the broader Soviet policy towards the Middle East region.

Coworking spaces: thriving and organizing in the new world of work

The present Ph.D. thesis proposes three studies on coworking spaces to understand how they foster thriving and organizing in the new world of work. The first study maps and analyzes the thematic structure and evolution of the academic debate that has emerged around coworking spaces in recent years. In doing so, it conducts a science mapping analysis of 351 publications on coworking spaces to detect and visualize key themes in the literature and their co-occurrence with subthemes. The second study proposes an interpretive review of 98 publications from multiple disciplines to shed light on how coworking spaces emerge as sites of organizing for professionals who are not formally connected to one another. It suggests five dimensions that articulate coworking spaces as sites of organizing – ‘materiality,’ ‘temporality,’ ‘affect,’ ‘identity,’ and ‘formalization.’ This study aims to go beyond the community-related understanding of coworking that has characterized most scholarly attention, instead focusing on coworking spaces’ organizational character. The third study investigates what drives thriving at work for remote workers in coworking spaces. In doing so, it acknowledges the potential complex set of interrelationships underpinning thriving at work and mobilizes complexity theory and qualitative comparative analysis to uncover six different, yet equifinal, configurations of antecedents driving remote workers’ thriving in coworking spaces.