Signal Processing, Multimedia and Biomedical Instrumentation (SPMBI) - Current Activities

Signal Separation, Fusion and Information Retrieval (Speech and audio processing, blind source separation, seismic wavefield separation, neural networks, machine learning, etc)
We have made major breakthroughs in establishing complete mathematical proofs for nonlinear signal processing theory. This has enabled us to established non-linear models for optimal signal separation and information retrieval. One of the earlier successes is the development of a complete framework using statistical methodologies for signal fusion, separation of complex nonlinearly mixed signals and informational retrieval. This work has radically challenged conventional approaches and research for speech, audio and seismic analysis show outstanding performance. The work has significantly contributed to the fundamentals of nonlinear signal processing theory and paved the way for the research community to develop platforms for optimal signal processing.

Biometric Recognition (3D face recognition, Iris recognition, keystroke dynamics, palmprint, cancellable biometrics, multimodal biometrics, security, etc.)
We have developed multivariate nonlinear models for practical biometric recognition problems to reduce dimensions and increase accuracy. In sharp contrast to conventional thinking, the nature of biometrics data is both highly dynamic and nonlinear. Traditional linear and static approaches for authentication often results in sub-optimal performance. Our approach of mathematical based framework incorporating nonlinearity and dynamical behaviours, has led to a new paradigm for spatio-temporal pattern recognition problems. New solutions which incorporate security issues and 3D models using these approaches have resulted in significant performance improvements. In multi-dimensional face recognition problems, enhanced models have been developed to encode the dynamical and non-linear effects for complex face images.

Biomedical and Instrumentation (Breast cancer, colon cancer, non-invasive bio and cellular markers, non-destructive evaluation, sensor technologies, etc)
This area of work has a world-wide reputation in biomedical research and our expertise includes colon and breast cancer classification and management. Work in this area covers computational intelligence using neuro-fuzzy networks for non invasive biomarkers, various biomedical image acquisition and analysis methods. The Group is also engaged in the development of multi-modal sensing systems for biomedical instrumentation in collaboration with leading Gastroenterologists and the Regional Neuroscience Centre. Current research has the potential to revolutionise endoscopy for colorectal cancer screening and the assessment of Parkinson's disease symptoms respectively.

Collaborations

Academic: Queen Mary University of London, Sheffield University, Queen's University Belfast, Sheffield Hallam, Institute of Molecular Medicine USA, Northumbria University, Regional Hospitals, Technical University of Cartagena, IIT Kharagpur, Malaysia Multimedia University.
Industrial: Foster Findlay, QinetiQ, e-Therapeutics, Torridon, Nortel, Office of the Deputy Prime Minister.

Contact

Prof Satnam Dlay, email: s.s.dlay@ncl.ac.uk, Tel: +44 (0) 191 222 8356.