University of Bremen

About: University of Bremen is a education organization based out in Bremen, Germany. It is known for research contribution in the topics: Population & Glacial period. The organization has 14563 authors who have published 37279 publications receiving 970381 citations. The organization is also known as: Universität Bremen.

On Detecting Adversarial Perturbations

Abstract: Machine learning and deep learning in particular has advanced tremendously on perceptual tasks in recent years. However, it remains vulnerable against adversarial perturbations of the input that have been crafted specifically to fool the system while being quasi-imperceptible to a human. In this work, we propose to augment deep neural networks with a small "detector" subnetwork which is trained on the binary classification task of distinguishing genuine data from data containing adversarial perturbations. Our method is orthogonal to prior work on addressing adversarial perturbations, which has mostly focused on making the classification network itself more robust. We show empirically that adversarial perturbations can be detected surprisingly well even though they are quasi-imperceptible to humans. Moreover, while the detectors have been trained to detect only a specific adversary, they generalize to similar and weaker adversaries. In addition, we propose an adversarial attack that fools both the classifier and the detector and a novel training procedure for the detector that counteracts this attack.

Flame spray pyrolysis: An enabling technology for nanoparticles design and fabrication

Abstract: Combustion of appropriate precursor sprays in a flame spray pyrolysis (FSP) process is a highly promising and versatile technique for the rapid and scalable synthesis of nanostuctural materials with engineered functionalities. The technique was initially derived from the fundamentals of the well-established vapour-fed flame aerosols reactors that was widely practised for the manufacturing of simple commodity powders such as pigmentary titania, fumed silica, alumina, and even optical fibers. In the last 10 years however, FSP knowledge and technology was developed substantially and a wide range of new and complex products have been synthesised, attracting major industries in a diverse field of applications. Key innovations in FSP reactor engineering and precursor chemistry have enabled flexible designs of nanostructured loosely-agglomerated powders and particulate films of pure or mixed oxides and even pure metals and alloys. Unique material morphologies such as core–shell structures and nanorods are possible using this essentially one step and continuous FSP process. Finally, research challenges are discussed and an outlook on the next generation of engineered combustion-made materials is given.

Dynamic oxidation of gaseous mercury in the Arctic troposphere at polar sunrise.

Abstract: Gaseous elemental mercury (Hg0) is a globally distributed air toxin with a long atmospheric residence time. Any process that reduces its atmospheric lifetime increases its potential accumulation in the biosphere. Our data from Barrow, AK, at 71 degrees N show that rapid, photochemically driven oxidation of boundary-layer Hg0 after polar sunrise, probably by reactive halogens, creates a rapidly depositing species of oxidized gaseous mercury in the remote Arctic troposphere at concentrations in excess of 900 pg m(-3). This mercury accumulates in the snowpack during polar spring at an accelerated rate in a form that is bioavailable to bacteria and is released with snowmelt during the summer emergence of the Arctic ecosystem. Evidence suggests that this is a recent phenomenon that may be occurring throughout the earth's polar regions.

Mechanisms of face perception.

Abstract: Faces are among the most informative stimuli we ever perceive: Even a split-second glimpse of a person's face tells us his identity, sex, mood, age, race, and direction of attention. The specialness of face processing is acknowledged in the artificial vision community, where contests for face-recognition algorithms abound. Neurological evidence strongly implicates a dedicated machinery for face processing in the human brain to explain the double dissociability of face- and object-recognition deficits. Furthermore, recent evidence shows that macaques too have specialized neural machinery for processing faces. Here we propose a unifying hypothesis, deduced from computational, neurological, fMRI, and single-unit experiments: that what makes face processing special is that it is gated by an obligatory detection process. We clarify this idea in concrete algorithmic terms and show how it can explain a variety of phenomena associated with face processing.