Technical University of Berlin

About: Technical University of Berlin is a education organization based out in Berlin, Germany. It is known for research contribution in the topics: Quantum dot & Laser. The organization has 27292 authors who have published 59342 publications receiving 1414623 citations. The organization is also known as: Technische Universität Berlin & TU Berlin.

Interactive Perception: Leveraging Action in Perception and Perception in Action

Abstract: Recent approaches in robot perception follow the insight that perception is facilitated by interaction with the environment. These approaches are subsumed under the term Interactive Perception (IP). This view of perception provides the following benefits. First, interaction with the environment creates a rich sensory signal that would otherwise not be present. Second, knowledge of the regularity in the combined space of sensory data and action parameters facilitates the prediction and interpretation of the sensory signal. In this survey, we postulate this as a principle for robot perception and collect evidence in its support by analyzing and categorizing existing work in this area. We also provide an overview of the most important applications of IP. We close this survey by discussing remaining open questions. With this survey, we hope to help define the field of Interactive Perception and to provide a valuable resource for future research.

The IPCC AR5 guidance note on consistent treatment of uncertainties: a common approach across the working groups

Abstract: Evaluation and communication of the relative degree of certainty in assessment findings are key cross-cutting issues for the three Working Groups of the Intergovernmental Panel on Climate Change. A goal for the Fifth Assessment Report, which is currently under development, is the application of a common framework with associated calibrated uncertainty language that can be used to characterize findings of the assessment process. A guidance note for authors of the Fifth Assessment Report has been developed that describes this common approach and language, building upon the guidance employed in past Assessment Reports. Here, we introduce the main features of this guidance note, with a focus on how it has been designed for use by author teams. We also provide perspectives on considerations and challenges relevant to the application of this guidance in the contribution of each Working Group to the Fifth Assessment Report. Despite the wide spectrum of disciplines encompassed by the three Working Groups, we expect that the framework of the new uncertainties guidance will enable consistent communication of the degree of certainty in their policy-relevant assessment findings.

Efficient and Accurate Lp-Norm Multiple Kernel Learning

Abstract: Learning linear combinations of multiple kernels is an appealing strategy when the right choice of features is unknown. Previous approaches to multiple kernel learning (MKL) promote sparse kernel combinations to support interpretability. Unfortunately, l1-norm MKL is hardly observed to outperform trivial baselines in practical applications. To allow for robust kernel mixtures, we generalize MKL to arbitrary lp-norms. We devise new insights on the connection between several existing MKL formulations and develop two efficient interleaved optimization strategies for arbitrary p > 1. Empirically, we demonstrate that the interleaved optimization strategies are much faster compared to the traditionally used wrapper approaches. Finally, we apply lp-norm MKL to real-world problems from computational biology, showing that non-sparse MKL achieves accuracies that go beyond the state-of-the-art.

The 2020 motile active matter roadmap

Abstract: Activity and autonomous motion are fundamental in living and engineering systems. This has stimulated the new field of 'active matter' in recent years, which focuses on the physical aspects of propulsion mechanisms, and on motility-induced emergent collective behavior of a larger number of identical agents. The scale of agents ranges from nanomotors and microswimmers, to cells, fish, birds, and people. Inspired by biological microswimmers, various designs of autonomous synthetic nano- and micromachines have been proposed. Such machines provide the basis for multifunctional, highly responsive, intelligent (artificial) active materials, which exhibit emergent behavior and the ability to perform tasks in response to external stimuli. A major challenge for understanding and designing active matter is their inherent nonequilibrium nature due to persistent energy consumption, which invalidates equilibrium concepts such as free energy, detailed balance, and time-reversal symmetry. Unraveling, predicting, and controlling the behavior of active matter is a truly interdisciplinary endeavor at the interface of biology, chemistry, ecology, engineering, mathematics, and physics. The vast complexity of phenomena and mechanisms involved in the self-organization and dynamics of motile active matter comprises a major challenge. Hence, to advance, and eventually reach a comprehensive understanding, this important research area requires a concerted, synergetic approach of the various disciplines. The 2020 motile active matter roadmap of Journal of Physics: Condensed Matter addresses the current state of the art of the field and provides guidance for both students as well as established scientists in their efforts to advance this fascinating area.

The Description-Experience Gap in Risky Choice: The Role of Sample Size and Experienced Probabilities

Abstract: Risky prospects come in different forms. Sometimes options are presented with convenient descriptions summarizing outcomes and their respective likelihoods. People can thus make decisions from description. In other cases people must call on their encounters with such prospects, making decisions from experience. Recent studies report a systematic and large description-experience gap. One key determinant of this gap is people`s tendency to rely on small samples resulting in substantial sampling error. Here we examine whether this gap exists even when people draw on large samples. Although smaller, the gap persists. We use the choices of the present and previous studies to test a large set of candidate strategies that model decisions from experience, including 12 heuristics, two associative-learning models and the two-stage model of cumulative prospect theory. This model analysis suggests-as one explanation for the remaining description-experience gap in large samples-that people treat probabilities differently in both types of decisions. Copyright (c) 2008 John Wiley & Sons, Ltd.