Research

Automatic Speech Recognition

I worked on automatic speech recognition (ASR) as part of the Natural Speech Technology project, where I was involved primarily on RNN/LSTM language modelling for multi-genre broadcast ASR.

Visual Speech Synthesis

My PhD work at The University of Manchester was on improved coarticulation modelling using long-range contexts as well as the joint modelling of audio and visual data using Gaussian processes (GPs). My PhD thesis can be accessed on the following link.

Facial Analysis and Animation

I spent a couple of years in industry and worked on really exciting projects at Image Metrics. I was involved both on the professional facial animation pipeline with contributions to improving Image Metrics proprietary algorithms for facial mapping, as well as on the latest consumer technologies involving the release of mobile apps for virtual try-ons, using augmented reality.