LIVE

LIVE is an Integrated Project partially funded under the European Union's IST Sixth Framework Programme. The LIVE project promotes a new, third market segment in the digital interactive television sector that does not exist today: intelligent television programming and services. The LIVE solution radically improves on linear approaches to TV broadcasting of live events by providing digital technologies andcontent formats that enable viewers to shape their own personal and highly interactive viewing experience.

To be more specific, the ORF used real-time consumer feedback in combination with sophisticted recommender reseach tools to produce 4 LIVE channels to cover multible events during the 2008 Olympic Games in Athens. I worked as a work package leader for metadata generation on preparing and indexing archive content and on creating realtime automatic as well as manual annotation tools to enrich the meta data covering the event in real time. This metadata was used as input to the LIVE recommender system to suggest useful achive clips to the LIVE production team ready for additional play-out. 

AGMA

AGMA is the short name for Automatic Generation of Meta data of Audio-visual content in context of MPEG-7.  The goal of the 3 year project funded by the BMBF from 2001 to 2003 was to develop new methods for automatic content analysis and to integrate these methods in a MPEG-7 compliant retrieval system, now known as the IFINDER solution. The components of the project covers speech recognition, music recognition, object segmentation and face recognition. I lead the project from 2002 to 2003 as a site manager and worked on image analysis and system engineering.

Read more ...

VIRTUAL HUMAN

The BMBF-Project Virtual Human (2002-2006) aims at creating realistic computer avatars with whom human users can easily interact and communicate. 

The major task for Fraunhofer IAIS/IMK was to create virtual humans by cloning real humans, to animate them (fullbody and facial movements) and build senses (speech recognition and computer vision) to interact with them. I lead the project as a site manager from 2004-2006 and worked on person localization, gesture recognition and full-body motion capture. 

ICE ROUTES

The European union research project ICE ROUTES was a project part-funded under the European Union's Fourth Framework programme and overseen by DGVII-E, the marine transport policy unit. The overall goal of this research project was to demonstrate the feasibility of an automatic ice routing tool for safer and more efficient ship transport in ice-covered waters. Research was focued on automatic tools for analyzing RADARSAT SAR imagery, passive microwave, weather and geographic data to generate ice maps used as input into ship routing systems. I worked as a scientist at the Ruhr-Uni Bochum, Institut für Neuroninformatik, Lehrstuhl Prof. von der Malsburg together with Bernd Fritzke on automatic sea ice classification with neural networks.

Read more...

iFinderSDK

The iFinderSDK is a C++-SDK for multimedia indexing developed at the Fraunhofer Institute IAIS. It contains modules for speech and video indexing, face and object recognition and has interfaces to databases, middleware and metadata formats such as MPEG-7.  It is used in many multimedia indexing projects at the institute. I coordinated the development team until 2009.  

FLAVOR

FLAVOR is a C++ Software framework for Elastic Graph Matching developed at the Institut für Neuroinformatik at the Ruhr-University Bochum. It has been used for rapid prototyping of face and object recognition algorithms in various research studies in the institute.