ETRO-DSSP (Laboratory for Digital Speech and Audio Processing) is a research cell of the Department of Electronics and Information Processing ETRO in the faculty of Applied Sciences of the Vrije Universiteit Brussel (VUB). It is rooted in seminal research on speech signal analysis and synthesis that was initiated in 1980. Since that time the laboratory has acquired international recognition for its achievements, especially in the field of speech modification systems where we have introduced the well-known time-scaling algorithm WSOLA, as well as several contributions on text to speech synthesis, voice modification, and many other subjects.
In the last decade, we continued to strengthen and expand our core competences in the area of speech processing and we entered the neighbouring domains of audio processing and modelling with a special interest in musical signals and animal sounds. We are currently active in both the fields of speech technology and audio processing with several research projects in cooperation with scientific institutes as well as commercial companies. Our main current projects are in the areas of speech synthesis (text to speech synthesis, expressive synthesis, audiovisual speech synthesis (3D and 2D photorealistic)), speech recognition (expressiveness and robustness), and speech modification (speech enhancement and noise reduction, time scaling, dialog replacement and karaoke).
Together with the IRIS Research Group of ETRO, we have begun to structure our common efforts in the field of audiovisual signal processing as a result of our increased involvement in research and development programs on multi-modal signal processing systems.