Key research themes
1. How can time-frequency analysis and adaptive signal processing improve biomedical signal interpretation and health monitoring?
This research theme explores advanced signal processing techniques such as time-frequency representations (e.g., Hilbert-Huang transform, wavelet transforms, Stockwell transform), adaptive filtering, and deep learning for robust interpretation of complex, non-stationary biomedical signals. The focus is on enhancing the extraction of physiological information (e.g., respiratory rate, ECG components) from noisy and time-varying data, which is critical for accurate diagnostics and continuous health monitoring, especially in wearable sensor applications.
2. What novel mathematical frameworks and filter designs advance signal decomposition and feature extraction in multidimensional and nonstationary signal environments?
Research under this theme tackles the theoretical foundations and algorithmic implementations of signal decomposition methods using analytic signals, multidimensional generalizations, higher-dimensional Hilbert transforms, and symmetric variable-length filtering. Such techniques resolve challenges of nonstationarity, mode mixing, and high-dimensionality, enabling improved feature extraction and noise resilience in image processing and other complex signal processing tasks.
3. How are practical implementations and software platforms advancing the deployment of sophisticated digital signal processing systems?
This stream focuses on the design, implementation, and hardware-software integration aspects of digital signal processing systems using contemporary programming environments and hardware architectures. It underscores the role of graphical programming (e.g., LabVIEW), fixed-point and floating-point arithmetic considerations, filter design toolkits, and DSP processor features that facilitate real-time, high-performance signal processing in industrial and academic applications.