CAIR Deep Learning performs basic and applied research in machine learning, with an emphasis on the theory and application of deep learning. Theoretical themes: (1) Generalisation in deep learning; (2) Transformation invariance; (3) Interpretability. Building on a strong track record in the application of machine learning to multilingual speech processing, our application domains are diverse, with the two main application themes (4) multilingual speech processing and (5) geospatial applications.
Applications are developed in collaboration with domain partners from industry and government, with the research group itself focussing on the underpinning techniques and tools. CAIR Deep Learning is housed in Multilingual Speech Technologies (MuST), a research group within the Faculty of Engineering of North-West University.