elib
DLR-Header
DLR-Logo -> http://www.dlr.de
DLR Portal Home | Imprint | Privacy Policy | Contact | Deutsch
Fontsize: [-] Text [+]

Discriminative regularization of the latent manifold of variational auto-encoders

Kossyk, Ingo and Marton, Zoltan-Csaba (2019) Discriminative regularization of the latent manifold of variational auto-encoders. Journal of Visual Communication and Image Representation, 61, pp. 121-129. Elsevier. DOI: 10.1016/j.jvcir.2019.03.008 ISSN 1047-3203

[img] PDF - Registered users only until 1 June 2021 - Postprint version (accepted manuscript)
3MB

Official URL: http://dx.doi.org/10.1016/j.jvcir.2019.03.008

Abstract

We present an approach on training classifiers or regressors using the latent embedding of variational auto-encoders (VAE), an unsupervised deep learning method, as features. Usually VAEs are trained using unlabeled data and independently from the classifier, whereas we investigate and analyze the performance of a classifier or regressor that is trained jointly with the variational deep network. We found that models trained this way can improve the embedding s.t. to increase classification performance, and also can be used for semi-supervised learning, building up the information extracting latent representation in an incremental fashion. The model was tested on two widely known computer vision benchmarks, and its generalization power was evaluated on an independent dataset. Additionally, generally applicable statistical methods are presented for evaluating similarly performing classifiers, and used to quantify the performance increase. The general applicability and ease-of-use of deep learning approaches allows for a wide applicability of the method.

Item URL in elib:https://elib.dlr.de/128125/
Document Type:Article
Title:Discriminative regularization of the latent manifold of variational auto-encoders
Authors:
AuthorsInstitution or Email of AuthorsAuthors ORCID iD
Kossyk, IngoIngo.Kossyk (at) dlr.deUNSPECIFIED
Marton, Zoltan-CsabaZoltan.Marton (at) dlr.dehttps://orcid.org/0000-0002-3035-493X
Date:May 2019
Journal or Publication Title:Journal of Visual Communication and Image Representation
Refereed publication:Yes
Open Access:No
Gold Open Access:No
In SCOPUS:Yes
In ISI Web of Science:Yes
Volume:61
DOI :10.1016/j.jvcir.2019.03.008
Page Range:pp. 121-129
Publisher:Elsevier
ISSN:1047-3203
Status:Published
Keywords:Variational auto-encoder Regularization Knowledge representation Perceptual data compaction Semi-supervised learning Statistical performance analysis
HGF - Research field:Aeronautics, Space and Transport
HGF - Program:Space
HGF - Program Themes:Space Technology
DLR - Research area:Raumfahrt
DLR - Program:R SY - Technik für Raumfahrtsysteme
DLR - Research theme (Project):R - Vorhaben Multisensorielle Weltmodellierung, R - Vorhaben Intuitive Mensch-Roboter Schnittstelle
Location: Oberpfaffenhofen
Institutes and Institutions:Institute of Robotics and Mechatronics (since 2013) > Perception and Cognition
Institute of Robotics and Mechatronics (since 2013) > Cognitive Robotics
Deposited By: Marton, Dr. Zoltan-Csaba
Deposited On:28 Jun 2019 13:00
Last Modified:28 Jun 2019 13:00

Repository Staff Only: item control page

Browse
Search
Help & Contact
Information
electronic library is running on EPrints 3.3.12
Copyright © 2008-2017 German Aerospace Center (DLR). All rights reserved.