See Symphony Automated Identification and Visualization of Emotions in Music

摘要

It is possible to automatically identify the emotions portrayed by music anddisplay those same emotions visually through the use of computer hardware and software. This thesis deals with a wide range of topics. Most of the areas are not computer science related, but they must be covered as a foundation for the work that is accomplished in this thesis and for the work that will follow. Chapter 2 is the literature review of topics relevant to this thesis. These topics include: properties of music, music and emotion, instruments and pitch, psychology of color, deafness, and previously related work in the field of manually and automatically generated visual music. Chapter 3 deals with the method used to show that the hypothesis is possible. Chapter 3 covers the following: instrument and music used, translation from music to emotion to animation, and software implementation. The software section of Chapter 3 also describes how the software was tested with deaf and hearing individuals. The results of the methods are detailed in Chapter 4. Conclusions and suggestions for further research are discussed in Chapter 5. Appendices A, B, and C contain descriptions, cover letters, and questionnaires used in the testing of the software with deaf and hearing subjects. Appendix D contains messages from six hearing subjects involved in the testing of the software.