My primary research interest is in endowing machines with audio (music) and vision understanding skills, such as the ability to recognise the genre, the structure and the emotions conveyed by a music piece; the ability to detect, track, and recognise (audio-visual) objects, faces, and facial expressions; and the potential to characterise the behaviour of humans from audio, visual, and multi-modal data. Developing machines with audio and visual understanding capabilities has many potential applications in the era of data deluge. Some examples include the effective organisation and exploration of large audio, image, and video collections, the development of (personalised) multimedia recommendation services, the design of natural human-computer interaction interfaces, to name but a few.
[+] more
