176 citations · 336 across the 51 of their papers we have counts for
11 papers · 1 filter
On Invariance, Equivariance, Correlation and Convolution of Spherical Harmonic Representations for Scalar and Vectorial Data
Janis Keuper
The mathematical representations of data in the Spherical Harmonic (SH) domain has recently regained increasing interest in the machine learning community. This technical report gi…
Investigating Shifts in GAN Output-Distributions
Ricard Durall, Janis Keuper
A fundamental and still largely unsolved question in the context of Generative Adversarial Networks is whether they are truly able to capture the real data distribution and, conseq…
A Retail Product Categorisation Dataset
Febin Sebastian Elayanithottathil, Janis Keuper
Most eCommerce applications, like web-shops have millions of products. In this context, the identification of similar products is a common sub-task, which can be utilized in the im…
Combating Mode Collapse in GAN training: An Empirical Analysis using Hessian Eigenvalues
Ricard Durall, Avraam Chatzimichailidis, Peter Labus +1
Generative adversarial networks (GANs) provide state-of-the-art results in image generation. However, despite being so powerful, they still remain very challenging to train. This i…
PHS: A Toolbox for Parallel Hyperparameter Search
Peter Michael Habelitz, Janis Keuper
We introduce an open source python framework named PHS - Parallel Hyperparameter Search to enable hyperparameter optimization on numerous compute instances of any arbitrary python…
Scalable Hyperparameter Optimization with Lazy Gaussian Processes
Raju Ram, Sabine Müller, Franz-Josef Pfreundt +2
Most machine learning methods require careful selection of hyper-parameters in order to train a high performing model with good generalization abilities. Hence, several automatic s…
Unmasking DeepFakes with simple Features
Ricard Durall, Margret Keuper, Franz-Josef Pfreundt +1
Deep generative models have recently achieved impressive results for many real-world applications, successfully generating high-resolution and diverse samples from complex datasets…
GradVis: Visualization and Second Order Analysis of Optimization Surfaces during the Training of Deep Neural Networks
Avraam Chatzimichailidis, Franz-Josef Pfreundt, Nicolas R. Gauger +1
Current training methods for deep neural networks boil down to very high dimensional and non-convex optimization problems which are usually solved by a wide range of stochastic gra…
Stabilizing GANs with Soft Octave Convolutions
Ricard Durall, Franz-Josef Pfreundt, Janis Keuper
Motivated by recently published methods using frequency decompositions of convolutions (e.g. Octave Convolutions), we propose a novel convolution scheme to stabilize the training a…
Sparsity in Deep Neural Networks - An Empirical Investigation with TensorQuant
Dominik Marek Loroch, Franz-Josef Pfreundt, Norbert Wehn +1
Deep learning is finding its way into the embedded world with applications such as autonomous driving, smart sensors and aug- mented reality. However, the computation of deep neura…
Using GPI-2 for Distributed Memory Paralleliziation of the Caffe Toolbox to Speed up Deep Neural Network Training
Martin Kuehn, Janis Keuper, Franz-Josef Pfreundt
Deep Neural Network (DNN) are currently of great inter- est in research and application. The training of these net- works is a compute intensive and time consuming task. To reduce…