Our research areas

Audio coding for low delay applications using low delay filter banks or predictive coding

Separation of vocals and accompanying instruments using Deep Learning

Low-delay audio source separation, for example in the time domain using probabilistic optimization

Music Information Retrieval (MIR) for signal analysis and modeling, e. g., for the recognition of musical instruments and the modeling and synthesis of their sounds

Audio coding:
development of polyphase filter banks | ultra low delay audio coding | multirate signal processing | development of audio encoders, musical instrument coder

Image and 3D video processing:
development of algorithms for automated reconstruction of 3D models from image data (e. g. structure from motion, feature detection, texture mapping) | Development of methods for texture analysis and coding | Paramaterial object coding | Computer vision and compressed sensing

Our research approaches

Machine learning for audio and music processing

Probabilistic optimization algorithms for optimization and adaptation of separation algorithms, filter banks and deep neural networks

 

Research and study projectst3://page?uid=15932

Project Robotic - Bipedal robotst3://page?uid=15939

Show video from vimeo.com