Research

At the MuSAE Lab we are interested in acquiring a better understanding of the human multimodal signal processing steps and to use that knowledge to build biologically-inspired technologies across three themes:
 

Theme 1:

Anthropomorphic technologies

Representative Projects:

1. Distributed quality-aware multimedia communications
2. Neural correlates of natural and synthesized speech quality perception
3. Neurophysiological models for ‘Quality-of-Experience’ prediction
4. Psychoacoustic modelling for objective quality assessment of reverberant and dereverberated speech
5. Environment-robust speaker identification
6. Multimodal affective state modeling

Theme 2:

Health diagnostics and monitoring

Representative Projects:

1. Quality measurement and enhancement of biomedical signals (e.g., ECG, EEG, EMG)
2. EEG-based diagnosis of Alzheimer’s disease
3. Spectro-temporal signal analysis for cardio-pulmonary signal analysis
4. Assistive technologies for voiceless communications
5. Objective quantification of handwriting difficulties in school-aged children
6. Remote speech intelligibility monitoring of dysarthric speakers

Theme 3:

Human-machine interaction

Representative Projects:

1. Intelligent signal processing for body area networks
2. Hybrid brain-computer interfaces
3. Peripheral Autonomic Signals for Binary Control of Body-Machine Interfaces
4. BCI and eye-gaze based serious gaming
5. Adaptive multimodal HCIs
6. Robust voice-based human-machine interaction