scispace - formally typeset
Search or ask a question
Institution

Frankfurt Institute for Advanced Studies

FacilityFrankfurt am Main, Germany
About: Frankfurt Institute for Advanced Studies is a facility organization based out in Frankfurt am Main, Germany. It is known for research contribution in the topics: Quark–gluon plasma & Baryon. The organization has 798 authors who have published 2733 publications receiving 82799 citations. The organization is also known as: FIAS.


Papers
More filters
Proceedings Article
01 Jan 2017
TL;DR: In this paper, a two time-scale update rule (TTUR) was proposed for training GANs with stochastic gradient descent on arbitrary GAN loss functions, which has an individual learning rate for both the discriminator and the generator.
Abstract: Generative Adversarial Networks (GANs) excel at creating realistic images with complex models for which maximum likelihood is infeasible. However, the convergence of GAN training has still not been proved. We propose a two time-scale update rule (TTUR) for training GANs with stochastic gradient descent on arbitrary GAN loss functions. TTUR has an individual learning rate for both the discriminator and the generator. Using the theory of stochastic approximation, we prove that the TTUR converges under mild assumptions to a stationary local Nash equilibrium. The convergence carries over to the popular Adam optimization, for which we prove that it follows the dynamics of a heavy ball with friction and thus prefers flat minima in the objective landscape. For the evaluation of the performance of GANs at image generation, we introduce the `Frechet Inception Distance'' (FID) which captures the similarity of generated images to real ones better than the Inception Score. In experiments, TTUR improves learning for DCGANs and Improved Wasserstein GANs (WGAN-GP) outperforming conventional GAN training on CelebA, CIFAR-10, SVHN, LSUN Bedrooms, and the One Billion Word Benchmark.

3,755 citations

Journal ArticleDOI
05 Oct 2006-Neuron
TL;DR: Evidence that certain brain disorders, such as schizophrenia, epilepsy, autism, Alzheimer's disease, and Parkinson's are associated with abnormal neural synchronization is reviewed to suggest close correlations between abnormalities in neuronal synchronization and cognitive dysfunctions.

1,956 citations

Journal ArticleDOI
TL;DR: Dysfunctional oscillations may arise owing to anomalies in the brain's rhythm-generating networks of GABA (γ-aminobutyric acid) interneurons and in cortico-cortical connections.
Abstract: Converging evidence from electrophysiological, physiological and anatomical studies suggests that abnormalities in the synchronized oscillatory activity of neurons may have a central role in the pathophysiology of schizophrenia. Neural oscillations are a fundamental mechanism for the establishment of precise temporal relationships between neuronal responses that are in turn relevant for memory, perception and consciousness. In patients with schizophrenia, the synchronization of beta- and gamma-band activity is abnormal, suggesting a crucial role for dysfunctional oscillations in the generation of the cognitive deficits and other symptoms of the disorder. Dysfunctional oscillations may arise owing to anomalies in the brain's rhythm-generating networks of GABA (gamma-aminobutyric acid) interneurons and in cortico-cortical connections.

1,813 citations

Journal ArticleDOI
15 Jun 2007-Science
TL;DR: It is proposed that the pattern of synchronization flexibly determines thepattern of neuronal interactions, and that the mutual influence among neuronal groups depends on the phase relation between rhythmic activities within the groups.
Abstract: Brain processing depends on the interactions between neuronal groups. Those interactions are governed by the pattern of anatomical connections and by yet unknown mechanisms that modulate the effective strength of a given connection. We found that the mutual influence among neuronal groups depends on the phase relation between rhythmic activities within the groups. Phase relations supporting interactions between the groups preceded those interactions by a few milliseconds, consistent with a mechanistic role. These effects were specific in time, frequency, and space, and we therefore propose that the pattern of synchronization flexibly determines the pattern of neuronal interactions.

1,327 citations

Journal ArticleDOI
TL;DR: Evidence is reviewed suggesting that the resulting rhythmic network inhibition interacts with excitatory input to pyramidal cells such that the more excited cells fire earlier in the gamma cycle, enabling transmission and read out of amplitude information within a single gamma cycle without requiring rate integration.

1,089 citations


Authors

Showing all 809 results

NameH-indexPapersCitations
Wolf Singer12458072591
Peter Braun-Munzinger10052734108
R. Stock9642934877
G. Kozlov9033936161
Luciano Rezzolla9039426159
Walter Greiner84128251857
Igor Pshenichnov8336222699
Xiaofeng Zhu80106228158
Mikolaj Krzewicki7728418908
Ivan Kisel7538918330
David Edmund Johannes Linden7436118787
David Michael Rohr7121715111
Sergey Gorbunov7125815638
M. Bach7112314661
Miklos Gyulassy6935819140
Network Information
Related Institutions (5)
Brookhaven National Laboratory
39.4K papers, 1.7M citations

87% related

Los Alamos National Laboratory
74.6K papers, 2.9M citations

85% related

CERN
47.1K papers, 1.7M citations

85% related

Max Planck Society
406.2K papers, 19.5M citations

85% related

Argonne National Laboratory
64.3K papers, 2.4M citations

84% related

Performance
Metrics
No. of papers from the Institution in previous years
YearPapers
202312
202224
2021172
2020155
2019172
2018219