This talk will be about Corti's journey on implementing real-time Machine Learning models directly on speech signals and utilise this information for decision-making. We will also summarise our research endeavours within the domain of unsupervised learning.
• Speech is hard and far from solved
• Deep learning on an unsupervised domain has many gains
• In order to implement models within the state-of- the-art areas of machine learning one need to stay on top of research