The Optical Neume Recognition Project: A tool to investigate early staff-less music notation

We invite applications for a funded PhD student to work on a project entitled "Optical Neume Recognition" at the School of Computer Science, University of Birmingham, UK. This is a multi-national project with partners in Canada, Germany and the UK.

This studentship is funded by the Canadian SSHRC (Social Sciences and Humanities Research Council) for the first two years and the School of Computer Science at the University of Birmingham for the third year.

Aim
The goal of the Optical Neume Recognition Project is to develop and apply software to elicit musical directives from graphic symbols (neumes) found in the earliest Western musical notation (10th century and later). Given that individual medieval chant manuscripts could contain anywhere from 100,000 to 250,000 recurrences of approximately 20+ discrete neume types, the proposed software will allow musicologists to study these signs more quickly and accurately and, most importantly, in far greater depth than ever before. This software will be indispensable when comparing scribal traditions, isolating differences in chant melodies transmitted in various neumed manuscripts, and answering questions about: the meaning and uses of particular neumes, various types of neume-text relationships, and the scribal process itself, such as the use of writing materials and kinds of working methods a scribe might use.

To achieve these goals, advances in historical document analysis technology will be required in the areas of specialised noise reduction, binarisation, layout analysis, feature extraction, colour separation, handwriting analysis, symbol recognition, word spotting and text to script registration. Since the target software is intended to provide interactive assistance to musicologists working around the world, high quality interactive graphical web interfaces to the tools produced will be required. The successful candidate will conduct research on a one or more of these problems to assist in developing the advances necessary as part of his or her PhD.

Candidate
The successful candidate is expected to have a good undergraduate and/or masters degree in computer science or a related area. She or he must have outstanding logical and analytical skills. Good communication skills (both oral and written), people skills, the ability to work independently as well as as a team member, initiative and pro-activity are essential. Experience with one or more of the areas of image analysis (particularly document image analysis), machine learning, pattern recognition and web application development is desirable. Candidates should have excellent programming skills. An interest in or some knowledge of music and music notation would be advantageous in order to understand the full context of the project.