University of Freiburg

About: University of Freiburg is a education organization based out in Freiburg, Baden-Württemberg, Germany. It is known for research contribution in the topics: Population & Transplantation. The organization has 41992 authors who have published 77296 publications receiving 2896269 citations. The organization is also known as: alberto-ludoviciana & Albert-Ludwigs-Universität Freiburg.

Inverting Visual Representations with Convolutional Networks

Abstract: Feature representations, both hand-designed and learned ones, are often hard to analyze and interpret, even when they are extracted from visual data. We propose a new approach to study image representations by inverting them with an up-convolutional neural network. We apply the method to shallow representations (HOG, SIFT, LBP), as well as to deep networks. For shallow representations our approach provides significantly better reconstructions than existing methods, revealing that there is surprisingly rich information contained in these features. Inverting a deep network trained on ImageNet provides several insights into the properties of the feature representation learned by the network. Most strikingly, the colors and the rough contours of an image can be reconstructed from activations in higher network layers and even from the predicted class probabilities.

Asynchronous GABA release generates long-lasting inhibition at a hippocampal interneuron–principal neuron synapse

Abstract: Hippocampal GABAergic interneurons show diverse molecular and morphological properties. The functional significance of this diversity for information processing is poorly understood. Here we show that cholecystokinin (CCK)-expressing interneurons in rat dentate gyrus release GABA in a highly asynchronous manner, in contrast to parvalbumin (PV) interneurons. With a gamma-frequency burst of ten action potentials, the ratio of asynchronous to synchronous release is 3:1 in CCK interneurons but is 1:5 in parvalbumin interneurons. N-type channels trigger synchronous and asynchronous release in CCK interneuron synapses, whereas P/Q-type Ca(2+) channels mediate release at PV interneuron synapses. Effects of Ca(2+) chelators suggest that both a long-lasting presynaptic Ca(2+) transient and a large distance between Ca(2+) source and sensor of exocytosis contribute to the higher ratio of asynchronous to synchronous release in CCK interneuron synapses. Asynchronous release occurs at physiological temperature and with behaviorally relevant stimulation patterns, thus generating long-lasting inhibition in the brain.

Innate immune memory in the brain shapes neurological disease hallmarks

Abstract: Innate immune memory is a vital mechanism of myeloid cell plasticity that occurs in response to environmental stimuli and alters subsequent immune responses. Two types of immunological imprinting can be distinguished—training and tolerance. These are epigenetically mediated and enhance or suppress subsequent inflammation, respectively. Whether immune memory occurs in tissue-resident macrophages in vivo and how it may affect pathology remains largely unknown. Here we demonstrate that peripherally applied inflammatory stimuli induce acute immune training and tolerance in the brain and lead to differential epigenetic reprogramming of brain-resident macrophages (microglia) that persists for at least six months. Strikingly, in a mouse model of Alzheimer’s pathology, immune training exacerbates cerebral β-amyloidosis and immune tolerance alleviates it; similarly, peripheral immune stimulation modifies pathological features after stroke. Our results identify immune memory in the brain as an important modifier of neuropathology.

Embed to Control: A Locally Linear Latent Dynamics Model for Control from Raw Images

Abstract: We introduce Embed to Control (E2C), a method for model learning and control of non-linear dynamical systems from raw pixel images. E2C consists of a deep generative model, belonging to the family of variational autoencoders, that learns to generate image trajectories from a latent space in which the dynamics is constrained to be locally linear. Our model is derived directly from an optimal control formulation in latent space, supports long-term prediction of image sequences and exhibits strong performance on a variety of complex control problems.

Learning to Estimate 3D Hand Pose from Single RGB Images

Abstract: Low-cost consumer depth cameras and deep learning have enabled reasonable 3D hand pose estimation from single depth images. In this paper, we present an approach that estimates 3D hand pose from regular RGB images. This task has far more ambiguities due to the missing depth information. To this end, we propose a deep network that learns a network-implicit 3D articulation prior. Together with detected keypoints in the images, this network yields good estimates of the 3D pose. We introduce a large scale 3D hand pose dataset based on synthetic hand models for training the involved networks. Experiments on a variety of test sets, including one on sign language recognition, demonstrate the feasibility of 3D hand pose estimation on single color images.