Research
Eva's primary area of interest is in digital speech and audio signal processing: 3D audio recording and reproduction, microphone array design and processing, room acoustics modeling, psychoacoustics (especially in relation to spatial perception), and (real-time) audio synthesis techniques.
Her undergraduate honours thesis studied using physical modeling techniques to jointly synthesise the audio and visual deformations resulting from virtual object interactions. Applicable to animation industries and gaming, the research aims to enhance virtual environments by accurately modeling the physical phenomenon.
Commencing her PhD candidature in 2004, Eva's research topic investigates using spatial information and new microphone array recording techniques for segmenting and annotating meeting speech recordings in a semantically meaningful manner. The idea is to find a means to automatically annotate meeting audio but stepping away from the traditional techniques of purely using low-level audio features. Applications of this work would be in making large corpora of meeting audio easily accessible and searchable.
