PhD Studentship

PhD Project: Algorithms for Audio/Video Content Identification and Verification

A PhD studentship funded under the EPSRC Industrial CASE scheme is available within the School of Computing at the University of Leeds. The industrial partner is ZOO Digital Group plc, a UK-based software company that develops workflow automation systems for the film industry, with customers including Walt Disney Pictures and Television and Warner Bros. The student will be based in the School of Computing at Leeds, with supervision provided by Dr Mark Everingham at Leeds and Dr Stuart Green at ZOO Digital. The student will be expected to spend at least 3 months in placements at ZOO Digital over the course of the PhD.

The purpose of the project is to conduct fundamental research in computer vision, image and audio processing algorithms that will enable the creation of commercial systems for the fully automated and/or human-assisted analysis and identification of audio/video materials in the entertainment industry. The key project goal is to develop methods for efficient processing of audio-video streams to identify content and to detect variations, artefacts, errors and other anomalies. The application of this is in (a) the automated inspection of video programmes to assess quality and identify faults, (b) the tracking of content, and (c) the processing of archives to determine how multiple instances derived from the same content are related.

The developed algorithms will have the following objectives:

• Enable identification of audio-video content using digital fingerprinting techniques i.e. establish whether two sequences contain largely the same content or are clips from a common source. The methods should be robust to imaging variations such as different formats, aspect ratio, compression, subtitle overlays, etc. and to minor content variations such as deleted shots, censored audio, etc.
• Establish detailed correspondence between two related audio-video sources i.e. which parts of two sequences are in common and where there are differences e.g. deleted shots, retimed sequences or censored content. Again, the sources may differ by format, compression, etc. such that differences are in content rather than formatting or quality.
• Assess the quality of a video stream in terms of fidelity to a master of known high quality. In this case the evaluation of differences assumes the same content and is in terms of perceptual difference i.e. are there qualitative differences which are salient or undesirable to a human observer.
  

Value: The studentship is funded for 3.5 years and covers fees and maintenance at the standard EPSRC rate (currently £13,590 per annum).

This studentship is funded by an EPSRC industrial CASE award and will start as soon as the position is filled, no later than 1 April 2012. The successful candidate should fulfil the eligibility criteria for EPSRC funding through UK residency status and the studentship is therefore only open to UK students.

Entry requirements: Applicants should have or expect to obtain a first class or good 2.1 honours degree in mathematics, computer science or a relevant discipline.

Applicants should have strong mathematical skills and excellent C/C++ and/or MATLAB programming ability. An interest in computer vision, machine learning and/or signal processing is essential, and proven experience in one of these areas would be a distinct advantage.

Supervisor: Dr Mark Everingham

Application deadline: Open until filled

Further information: To discuss this project further informally, please contact Dr Mark Everingham, m.everingham@leeds.ac.uk

Please note, all applications should be sent to the School's Student Support Office, not Dr Mark Everingham.

How to apply: Formal applications for research degree study should be made on-line through the University's website. Please state clearly on the funding section of the application form that you wish to be considered for this scholarship. If you have published research papers, please list these in the 'Additional Information' section.

Please upload all the documents required as soon as possible. Scanned copies are acceptable though you will need to provide originals or certified copies at registration. These will include your degree certificate(s), transcripts of marks achieved in previous degrees, plus evidence of English language qualifications if your first language is not English and you do not hold a degree from an English-speaking country. Please do not provide school certificates, or non-academic certificates.

Please note, if you intend to send academic references we can only accept them if they are on official letter headed paper and contain an original signature and stamp; they must arrive in sealed envelopes. Alternatively, the School will contact your named academic referees directly.

If you require any further information please contact the Graduate School Office, e: phd@engineering.leeds.ac.uk, t: +44 (0)113 343 8000.