Goldsmiths, Department of Computing

Dr. Michael A. Casey

CIS320 Final Year Computing Project 2006/2007

These ideas are to help you define your own project. Make an appointment with me to discuss any of these, or to suggest your own version.
  • Multimedia Database Interface

    The aim of this project is to import XML-Schema based descriptions of digital video and audio into a database and to use the database to retrieve video clips based on various SQL queries.


  • Multimedia Retrieval Engine

    The aim of this project is to develop highly optimised internal representations of media segments and their metadata for digital video and audio. An intelligent multimedia database retrieval interface will use the database engine to retrieve video clips based on various SQL queries.

    XML, SQL, Java

  • VST/DirextX Plugin for Sound Un-mixing

    Musical audio is composed of simultaneously sounding sources (instruments). Using advanced signal processing techniques we can undo some of the mixing of the instruments to recover individual parts (under some important constraints).

    The goal of this project is to take the existing signal processing code and integrate it into a professional music / digital audio editing environment.

    C++ in Windows, with a willingness develop using DirectX or the VST libraries.

  • The Intelligent Music Player

    Many people have large music collections stored on hard-drives often consisting of thousands of songs. The problem is to find the music that you want when you want it. In this project you will implement a system for retrieving music by "mood", "style" or "genre" based on a library of automatic music annotation tools developed at Goldsmiths.

    Good JAVA or C++ skills and a keen interest in any kind of music.

  • Automatic Music Generator

    Using ideas from Cellular Automata and Autonomous Agents research, this project is to generate music using a set of simple rules applied to MIDI instruments. The music can be in any style: e.g. classical, jazz, dance music or blues.

    Good programming skills. AI/Agents. knowledge of MIDI. Understanding of music.

  • Music structure analysis tool

    Musicologists spend many years learning the intricacies of musical scores in order to make comparative evaluations between composers, styles and pieces. This project aims to provide a tool for automatically visualizing the structure of a musical piece using simple schematic diagrams and graphics. This tool could be used for educational or research purposes.

    good JAVA programming skills. Interface design. Some understanding of music.