Research Themes

Research Themes

The QxLab group is interested in models and metrics for all varieties of quality of experience, with a primary focus on media signal processing applications. The group is made up of post-doctoral, PhD and masters candidates looking at a wide variety of QoE related topics.

QoE for Listening. Listening technologies covering audio, music and speech are the main focus of QxLab. This includes topics such as realtime monitoring of Voice over IP speech quality monitoring, predicting speech intelligibility for hearing aids and predicting the effects of different environmental noise on speech. Here the focus is on the impact of the technology (e.g. a service or system) on the Quality of Experience.

Virtual reality and augmented reality are providing new experiences where beyond the fidelity of the audio and video, the immersive aspect of the experience is important. For virtual reality, 3D spatial audio is required, i.e. if a person is speaking behind the listener, their voice needs to be perceived as originating from the correct point in space with respect to the listener.

QoE for Media. Media QoE looks at the interaction of different types of content on QoE. This includes how media that has audio and visual components are perceived, e.g. is the QoE for a music video perceived differently depending on whether it is accompanied by a video? We also investigate how compression influences QoE, exploring audio codecs and bitrates.

Maintaining fidelity of recorded music archives is necessary to secure our shared heritage of audio, music and sound archives. These archives capture the sounds of the past and can be faithfully digitised from a degraded storage medium. Early sound archival methods, including wax cylinders, gramophone and vinyl records, degrade with time and are easily damaged at playback. Restoration algorithms for noise suppression, can restore high fidelity signals but they rely understanding and predicting their effect on Quality of Experience.

QoE Management. For traditional webpages or web applications such as maps or games, network delay is the primary issue when it comes to a poor Quality of Experience. We are looking at how network delays are perceived as human time perception is not a simple linear scale. In order to design and implement and manage improved networks, better models of web QoE are required.

QoE fundamentals. Quality of Experience is a relatively new concept. The frameworks that have been developed to help us understand, quantify and predict QoE have been developed from a multimedia perspective. However, the learning and ideas can be transferred to other aspects of human-technology interactions and to the development of data-driven models to predict human satisfaction, e.g. quality of life prediction for stroke patients.

QoE Security. New crypto-systems for multi-tenant applications have a range of cryptographic technologies and protocols at their disposal. Designing systems that are sufficiently secure that are also aware of the experience (performance speech and usability) from both a system user and provider perspective.