Otto-von-Guericke-Universität Magdeburg

 
 
 
 
 
 
 
 
David Hübner

Dipl.-Inf. David Philippou-Hübner


Institut für Informations- und Kommunikationstechnik (IIKT)

Gebäude 03, Universitätsplatz 2, 39106, Magdeburg, G03-324

Lebenslauf

seit 10/2007 Wissenschaftlicher Mitarbeiter am Lehrstuhl Kognitive Systeme
01/2007 -
06/2007
Wiss. Hilfskraft am Lehrstuhl für Theoretische Biologie und Biologische Kybernetik, Uni Bielefeld
01/2007 Diplom in Informatik;
Titel der Arbeit: Rekurrente Neuronale Netze. Verhalten von IC-Units bei der Lösung von homogenen Basisgleichungen und Differentialgleichungen.
2001 -
2006
Studentische Hilfskraft in der AG Praktische Informatik,
Tutor für Algorithmen & Datenstrukturen (mit Unterbrechungen)
1999 Beginn des Studiums der Naturwissenschaftlichen Informatik an der Uni Bielefeld
1998 - 1999 Zivildienst in Bielefeld
1990 - 1998 Städtisches Gymnasium Gütersloh; Abschluss Abitur
1978 Geboren in Berlin-Wilmersdorf

 

Forschungsinteressen

Situations-/Nutzer-adaptive Systeme, Mensch-Maschine-Kommunikation, Künstliche Neuronale Netze und (Maschinelles) Lernen, Biologische Optimierungsstrategien

 

Mitgliedschaften

Gesellschaft für Informatik (GI)

 

Tutorials / Software

HTK Quick-Start Tutorial, the "Hello-World" for the Hidden Markov Model Toolkit (.pdf, .zip).
A compact practice-oriented summary of the HTK Tutorial (Chapter 3, The HTK Book by S. Young et al.). In addition to all example files the .zip folder contains shell scripts/java programs which simplify and automate the work with HTK.

The HTK Feature Extractor is a java tool build around the HTK commands HCopy and HList, which are used to extract various features from audio files, especially in the speech processing community. The program helps decreasing the preparatory steps in order to apply HCopy to a minimum. In particular the program automates repetitive tasks like the analyses of the data structure of a given speech corpus and subsequently the time consuming creation of script files. Further, it increases processing speed by distributing load among the single cores of your machine. The extracted features will be stored in a folder of your choice maintaining the original folder structure of the source corpus. At the same time the complete HTK functionality is preserved (htk_feature_extractor.zip).

 

Betreute Abschlussarbeiten

Rene Kallweit
Sprachsteuerung eines Roboters über einen Rasberry PI bzw. eine Arduino Plattform.
Masterarbeit, 2014

Khalid Qadri
Current Flow Analysis in a Model for Electrical Brain Stimulation.
Masterarbeit, 2012

Matthias Steglich
Interaktive Sprachsteuerung zur dynamischen Festlegung von Prozessabläufen.
Diplomarbeit, 2011

Friedrich Melchert
Optimierung von HTK-Routinen für den Betrieb auf einem Rechnercluster.
Diplomarbeit, 2011

Michal Rapczynski
Emotionserkennung aus Sprache mit Hilfe von Feed-Forward Netzen unter Betrachtung von prosodischen (globalen) und lokalen Sprachmerkmalen.
Bachelorarbeit, 2011 (Betreut zusammen mit S. Glüge)

Nam Quang Tran
Emotionserkennung aus Sprache im Valence-Arousal-Raum mittels Neuronalen Netzen und Support-Vektor-Maschinen.
Diplomarbeit, 2010

Michael Preuß
Aufbau eines sprachgesteuerten Univis - Auskunftsystems.
Diplomarbeit, 2008

 

Publikationen

2014 2013 2012 2011 2010 2009 2008   [bib-tex]

2014

Siegert I, Philippou-Hübner D, Hartmann K,  Böck R, Wendemuth A
Investigation of Speaker Group-Dependent Modelling for Recognition of Affective States from Speech.
Cognitive Computation, available online, article in Press, 2014.

2013

Siegert I, Hartmann K, Philippou-Hübner D, Wendemuth A
Human behaviour in HCI: Complex Emotion Detection through Sparse Speech Features.
Proceedings of the 4th Int. Workshop on Human Behavior Understanding (HBU), October 22, Barcelona, Spain, 2013.

Hartmann K, Siegert I, Philippou-Hübner D, Wendemuth A
Emotion-Detection in HCI: From Speech Features to Emotion Space.
Proceedings of the 12th IFAC/IFIP/IFORS/IEA Symposium on Analysis, Design, and Evaluation of Human-Machine Systems, August 11-15, Las Vegas, USA, 2013.

Vlasenko B, Philippou-Hübner D, Wendemuth A
Parameter optimization issues for cross-corpora emotion classification.
Proceedings of the 5th Int. Conference on Affective Computing and Intelligent Interaction (ACII 2013), September 2-5, Geneva, Italy, 2013.

2012

Philippou-Hübner D, Böck R, Wendemuth A
Applying the Speaking Rate in a Hierarchical Classifier for Emotion Recognition from Speech.
Proceedings of the 23. Konferenz Elektronische Sprachsignalverarbeitung (ESSV 2012), August 29-31, Cottbus, Germany, 2012, ISBN: 978-3-942710-81-7.

Siegert I, Böck R, Philippou-Hübner D, Wendemuth A
Investigation of Hierarchical Classification for Simultaneous Gender and Age Recognition.
Proceedings of the 23. Konferenz Elektronische Sprachsignalverarbeitung (ESSV 2012), August 29-31, Cottbus, Germany, 2012, ISBN: 978-3-942710-81-7.

Philippou-Hübner D, Vlasenko B, Böck R, Wendemuth A
The Performance of the Speaking Rate Parameter in Emotion Recognition from Speech.

Proceedings of 2012 IEEE International Conference on Multimedia & Expo (ICME2012), July 9-13, Melbourne, Australia, 2012.

2011

Siegert I, Böck R, Philippou-Hübner D, Vlasenko B, Wendemuth A
Appropriate Emotional Labeling of Non-Acted Speech using Basic Emotions, Geneva Emotion Wheel and Self Assessment Manikins.
Proceedings of the 2011 IEEE International Conference on Multimedia & Expo (ICME2011), July 11-15, Barcelona, Spain, 2011.

Vlasenko B, Philippou-Hübner D, Prylipko D, Böck R, Siegert I, Wendemuth A
Vowels Formants Analysis allows straightforward Detection of High Arousal Emotions.
Proceedings of the 2011 IEEE International Conference on Multimedia & Expo (ICME2011), July 11-15, Barcelona, Spain, 2011.

Vlasenko B, Prylipko D, Philippou-Hübner D, Andreas Wendemuth A
Vowels formants analysis allows straightforward detection of high arousal acted and spontaneous emotions.
Proceedings of the INTERSPEECH 2011, August 28-31, Florence, Italy, 2011.

2010

Hübner D, Vlasenko B, Grosser T, Wendemuth A
Determining Optimal Features for Emotion Recognition from Speech by applying an Evolutionary Algorithm.
Proceedings of Interspeech 2010, September 26-30, Makuhari, Japan, 2010.

Hübner D, Wendemuth A
Artificial Neural Networks - A Connectionistic Approach to Human Cognition.
To appear: Proceedings of the 1. International Conference on What makes Humans Human (WMHH 2010), March 26-27, Ulm, Germany, 2010.

Hübner D, Wendemuth A
HMM-Based Emotion Recognition from Speech with Sparse Features.
Proceedings of Interdisciplinary College 2010 (IK 2010), March, 12-19, Günne, Germany, 2010.

Grosser T, Elsafadi A, Hübner D, Böck R, Wendemuth A
Adaptive Beamforming in Speaker Diarization.
Proceedings of the Berlin Beamforming Conference 2010 (BeBeC2010), February 24-25., Berlin, Germany, 2010.

Böck R, Hübner D, Wendemuth A
Determining Optimal Signal Features and Parameters for HMM-Based Emotion Classification.
Proceedings of the 15. IEEE Mediterranean Electrotechnical Conference (MELECON 2010), April 26-28, Valetta, Malta, 2010.

2009

Hübner D, Böck R, Wendemuth A
Application of HMMs for the Recognition of Emotional Sequences in the Valence-Arousal Space.
Proceedings of the 20. Konferenz Elektronische Sprachsignalverarbeitung (ESSV 2009), September 21-23, Dresden, Germany, 2009, ISBN: 978-3-941298-31-6.

Böck R, Hübner D, Wendemuth A
On the Influences of Feature Extraction in Single Emotion Recognition in Naive vs. Acted Speech.
Proceedings of the 20. Konferenz Elektronische Sprachsignalverarbeitung (ESSV 2009), September 21-23, Dresden, Germany, 2009, ISBN: 978-3-941298-31-6.

Hübner D, Grosser T, Wendemuth A
Dynamic Bayesian Network and Hidden Markov Models for Modeling Situations.
Proceedings of Research Workshop on "Emotion-, Speech-, and Face Recognition with advanced classifiers". A. Wendemuth and H.-G. Meyer (eds.), Magdeburg 2008, ISBN: 978-3-940961-24-2.

Grosser T, Böck R, Hübner D, Wendemuth A
Survey of State-of-the-Art Speech Recognition Technology.
Proceedings of Research Workshop on "Emotion-, Speech-, and Face Recognition with advanced classifiers". A. Wendemuth and H.-G. Meyer (eds.), Magdeburg 2008, ISBN: 978-3-940961-24-2.

2008

Cruse H, Hübner D
Selforganizing memory: active learning of landmarks used for navigation.
Biological Cybernetics 99, 219-236

Makarov VA, Song Y, Velarde MG, Hübner D, Cruse H
Elements for a general memory structure: properties of recurrent neural networks used to form situation models.
Biological Cybernetics 98, 371-395.  



Letzte Änderung: 11.03.2015 - Ansprechpartner: Dipl.-Inf. David Philippou-Hübner