Michael Feld is working as a full-time researcher in the DFKI’s Automotive Group. His PhD topic is "A Speaker Classification Framework for Non-intrusive User Modeling on the Example of Personalized In-Car Services ". A major part of his work focuses on the topics of user modelling, in particular user adaptation and personalization, as well as speech-based classification. Further research interests include application architecturing, parallel programming and mobile devices. Michael joined DFKI in early 2005 as a student research assistant. After finishing his Diploma thesis in Computer Science at Saarland University on the topic of "Erzeugung von Sprecherklassifikationsmodulen für multiple Plattformen"

[Generation of Speaker Classification Modules for Mobile Platforms] in 2006, he became a full-time researcher in the Intelligent User Interfaces (IUI) department headed by Prof. Wolfgang WahlsterSince then, he has been involved in several projects working in key areas such as speaker classification (and more general machine learning tasks), personalization and mobile devices. The most notable work was done in projects COLLATE, SBC, NetShield and Carmina. He has also been part of the Chair of Prof. Wahlster, co-teaching and assisting with lectures and seminars.

In the field of speaker classification, Michael has been working closely together with Dr. Christian Müller since his FoPra (Bachelor) thesis. One of his main accomplishments was the design and development of a complete framework for all tasks related to speech-based classification supporting specialized algorithms, strict resource requirements and flexible deployment. His PhD topic follows closely in on this, using the architecture to explore new methods that produce even better classification results. The results are subsequently used in project Carmina and therefore close the circle on the application end.

In Carmina, Michael is responsible for the aspects related to personalization of the Human/Machine Interface (HMI; i.e. GUI, speech, tangible interfaces...). For this purpose, he is exploring new possibilities of adaptation and customization of in-car services to user and context parameters, thereby incorporating innovative methods of information acquisition and inference. A particular focus is on speech as an input source, but also taken into account are the passengers' tasks/goals, preferences, abilities, roles, and usage data.

Speaker Classification Platform

If you are looking for the SPEACLAP (SPEAker CLAssification Platform) application, you can download it here. You will also need the .NET Framework 4 and MySQL to run it properly. If you have questions on using the application, you can always This e-mail address is being protected from spambots. You need JavaScript enabled to view it me.

Publications

* published before joining Automotive IUI

2010

Michael Feld, Felix Burkhardt, Christian Müller (2010): Automatic Speaker Age and Gender Recognition in the Car for Tailoring Dialog and Mobile Services. In Proceedings of Interspeech (2010). [bib]

Christoph Endres, Michael Feld, Tim Schwartz, Christian Müller (2010): Cinematic Analysis of Automotive Personalization. In Proceedings of the 2nd International Workshop on Multimodal Interfaces for Automotive Applications (MIAA 2010) in conjunction with the ACM Conference on Intelligent User Interfaces IUI 2010. [bib]

Charl Heerden, Etienne Barnard, Marelie Davel, Christiaan Walt, Ewald Dyk, Michael Feld, Christian Müller (2010): Combining Regression and Classification Methods for Improving Automatic Speaker Age Recognition. In Proceedings of the The 35th International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2010). [bib]

Michael Feld, Tim Schwartz, Christian Müller (2010): Second Multimodal Interfaces for Automotive Applications (MIAA 2010). In Proceedings of the International Conference on Intelligent User Interfaces (IUI 2010). 441--442. [bib]

Michael Feld, Christoph Endres (2010): Sharing User and Context Models in Automotive HMI. In Adjunct Proceedings of the 2nd International Conference on Automotive User Interfaces and Interactive Vehicular Applications (AutomotiveUI 2010). 10. [bib]

Michael Feld, Tim Schwartz, Christian Müller (2010): This is me: Using Ambient Voice Patterns for In-Car Positioning. In Proceedings of the 1st International Conference on Ambient Intelligence. [bib]

2009

Michael Feld, Christian Müller (2009): An Integrated Development Environment for Speech-Based Classification. In Proceedings of the 13th International Conference ``Speech and Computer\\'\\' SPECOM 2009. 443 -- 447. [bib]

Michael Feld, Etienne Barnard, Christian Müller (2009): Multilingual Speaker Age Recognition:Regression Analyses on the Lwazi Corpus. In Proceedings of the eleventh biannual IEEE workshop on Automatic Speech Recognition and Understanding (ASRU). 534--539. [bib]

2008

*   Michael Feld, Christian Mueller (2008): Speaker Classification for Mobile Devices. In Proceedings of the 2nd IEEE International Interdisciplinary Conference on Portable Information Devices (Portable 2008). [bib]