Deep Learning - Groups

Harmonic Decompositions of Convolutional Networks

The dataset used in this paper is a collection of images of faces, each with a different expression.

Dataset
JSON

H-Fac: Memory-Efficient Optimization with Factorized Hamiltonian Descent

The dataset used in this study is a benchmark for evaluating the performance of memory-efficient optimizers.

Dataset
JSON

Detecting Out-of-Distribution Inputs in Deep Neural Networks

The dataset used in the paper is a low- and high-dimensional datasets and deep models.

Dataset
JSON

Subregular Complexity and Deep Learning

The dataset consists of six formal target languages defined for training and testing purposes. Each language is a subset of the alphabet {a, b, c, d}.

Dataset
JSON

DeepSig dataset

The DeepSig dataset is a collection of synthetic RF signals from 24 modulation classes.

Dataset
JSON

Membership-Invariant Subspace Training

Membership-Invariant Subspace Training (MIST) is a method for training classifiers that acts as a defense designed to specifically defend against black-box membership inference...

Dataset
JSON

Attacking Adversarial Attacks as A Defense

The dataset used in the paper is not explicitly described, but it is mentioned that the authors examined state-of-the-art attacks of various kinds.

Dataset
JSON

GEEN Dataset

The dataset used in this paper is a sample of (X 1, X 2,..., X k, X ∗) where X ∗ is a latent variable.

Dataset
JSON

CUBDL

Ultrasound beamforming datasets for deep learning

Dataset
JSON

Langevin algorithms for Markovian Neural Networks and Deep Stochastic control

The dataset used in the paper is a stochastic control problem, where the control is parametrized by a neural network calibrated by gradient descent.

Dataset
JSON

Clustered Data Distribution

The dataset consists of clusters with means µ(1),..., µ(k) and the examples in the j-th cluster are labeled by y(j). The clusters are generated as follows: we draw j ∼ U[0, 1]...

Dataset
JSON

Gradient Adversarial Training

The dataset used for gradient adversarial training of neural networks.

Dataset
JSON

Fully Automatic Liver Attenuation Estimation combining CNN Segmentation and M...

The ALARM method is a fully automated liver attenuation estimation method that combines deep convolutional neural network (DCNN) and morphological operations.