Machine Learning – and very recently deep learning – have produced state-of-the-art results across various domains that span data analytics, computer vision, face and emotion recognition, and signal analysis. Such techniques can be adjusted to work with various data including still images, video files, numerical data etc. Having this in mind, new techniques and algorithms can be developed to help analyse large amounts of video files that might include humans who interact with other humans or objects, and either describe the action or, with the help of experts, provide a diagnosis. In autism, machine learning could be used to develop methods of assessing verbal and nonverbal communication skills in a reliable and easy to implement way. Thus this technology has the potential to facilitate the development of innovative diagnostic tools for autism and other related developmental conditions.
If you are interested in finding out more about our research please contact us