SciTePress - Affiliation Profile

Research.Publish.Connect.

Title:
ISBN:
Year:
Acronym:
Subject:

Advanced Search Proceedings Search

If you're looking for an exact phrase use quotation marks on text fields.

Title:
Author:
Affiliation:
Subject:

Advanced Search Papers Search

If you're looking for an exact phrase use quotation marks on text fields.

Name:
Affiliation:
Country:
Conference:
Subject:

Advanced Search Authors Search

If you're looking for an exact phrase use quotation marks on text fields.

Name:
Country:
Subject:

Advanced Search Affiliations Search

If you're looking for an exact phrase use quotation marks on text fields.

Proceedings

Proceedings Search

Title:
ISBN:
Year:
Acronym:
Subject:

Advanced Search Proceedings Search

If you're looking for an exact phrase use quotation marks on text fields.

Papers

Papers Search

Title:
Author:
Affiliation:
Subject:

Advanced Search Papers Search

If you're looking for an exact phrase use quotation marks on text fields.

Authors

Authors Search

Name:
Affiliation:
Country:
Conference:
Subject:

Advanced Search Authors Search

If you're looking for an exact phrase use quotation marks on text fields.

Advanced Search

Affiliation

Institut für Mathematik und Informatik, Universität Greifswald, Walther-Rathenau-Straße 47, 17489 Greifswald, Germany

Context (please use complete words only)

Field

Sort by

3 Papers

Calculating the Credibility of Test Samples at Inference by a Layer-wise Activation Cluster Analysis of Convolutional Neural Networks

Topics: Convolutional Neural Networks (CNN); Clustering, Classification and Regression; Image Classification

Authors: Daniel Lehmann and Marc Ebner

Affiliation: Institut für Mathematik und Informatik, Universität Greifswald, Walther-Rathenau-Straße 47, 17489 Greifswald, Germany

Digital Object Identifier: 10.5220/0011274000003277

Proceedings of DeLTA 2022, ISBN: 978-989-758-584-5

Quick Abstract

A convolutional neural network model is able to achieve high classification performance on test samples at inference, as long as those samples are drawn from the same distribution as the samples used for model training. However, if a test sample is drawn from a different distribution, the performance of the model decreases drastically. Such a sample is typically referred to as an out-of-distribution (OOD) sample. Papernot and McDaniel (2018) propose a method, called Deep k-Nearest Neighbors (DkNN), to detect OOD samples by a credibility score. However, DkNN are slow at inference as they are based on a kNN search. To address this problem, we propose a detection method that uses clustering instead of a kNN search. We conducted experiments with different types of OOD samples for models trained on either MNIST, SVHN, or CIFAR10. Our experiments show that our method is significantly faster than DkNN, while achieving similar performance.

Subclass-based Undersampling for Class-imbalanced Image Classification

Topics: Machine Learning Technologies for Vision; Categorization and Scene Understanding

Authors: Daniel Lehmann and Marc Ebner

Affiliation: Institut für Mathematik und Informatik, Universität Greifswald Walther-Rathenau-Straße 47, 17489 Greifswald, Germany

Digital Object Identifier: 10.5220/0010841100003124

Proceedings of VISAPP 2022/VISIGRAPP, ISBN: 978-989-758-555-5

Quick Abstract

Image classification problems are often class-imbalanced in practice. Such a class imbalance can negatively affect the classification performance of CNN models. A State-of-the-Art (SOTA) approach to address this issue is to randomly undersample the majority class. However, random undersampling can result in an information loss because the randomly selected samples may not come from all distinct groups of samples of the class (subclasses). In this paper, we examine an alternative undersampling approach. Our method undersamples a class by selecting samples from all subclasses of the class. To identify the subclasses, we investigated if clustering of the high-level features of CNN models is a suitable approach. We conducted experiments on 2 real-world datasets. Their results show that our approach can outperform a) models trained on the imbalanced dataset and b) models trained using several SOTA methods addressing the class imbalance.

Are Image Patches Beneficial for Initializing Convolutional Neural Network Models?

Topics: Machine Learning Technologies for Vision; Deep Learning for Visual Understanding

Authors: Daniel Lehmann and Marc Ebner

Affiliation: Institut für Mathematik und Informatik, Universität Greifswald, Walther-Rathenau-Straße 47, 17489 Greifswald, Germany

Digital Object Identifier: 10.5220/0010206603460353

Proceedings of VISAPP 2021/VISIGRAPP, ISBN: 978-989-758-488-6

Quick Abstract

Before a neural network can be trained the network weights have to be initialized somehow. If a model is trained from scratch, current approaches for weight initialization are based on random values. In this work we examine another approach to initialize the weights of convolutional neural network models for image classification. Our approach relies on presetting the weights of convolutional layers based on information given in the training images. To initialize the weights of convolutional layers we use small patches extracted from the training images to preset the filters of the convolutional layers. Experiments conducted on the MNIST, CIFAR-10 and CIFAR-100 dataset show that using image patches for the network initialization performs similar to state-of-the-art initialization approaches. The advantage is that our approach is more robust with respect to the learning rate. When a suboptimal value for the learning rate is used for training, our approach performs slightly better than current approaches. As a result, information given in the training images seems to be useful for network initialization resulting in a more robust training process.