PiaNote: A Sight-Reading Program That Algorithmically Generates Music Based on Human Performance

<p>Sight-reading is the act of performing a piece of music at first sight. This can be a difficult task to master, because it requires extensive knowledge of music theory, practice, quick thinking, and most importantly, a wide variety of musical material. A musician can only effectively sight-read with a new piece of music. This not only requires many resources, but also musical pieces that are challenging while also within a player's abilities.</p> <p>This thesis presents PiaNote, a sight-reading web application for pianists that algorithmically generates music based on human performance. PiaNote's goal is to alleviate some of the hassles pianists face when sight-reading. PiaNote presents musicians with algorithmically generated pieces, ensuring that a musician never sees the same piece of music twice. PiaNote also monitors player performances in order to intelligently present music that is challenging, but within the player's abilities. As a result, PiaNote offers a sight-reading experience that is tailored to the player.</p> <p>On a broader level, this thesis explores different methods in effectively creating a sight-reading application. We evaluate PiaNote with a user study involving novice piano players. The players actively practice with PiaNote over three fifteen-minute sessions. At the end of the study, users are asked to determine whether PiaNote is an effective practice tool that improves both their confidence in sight-reading and their sight-reading abilities. Results suggest that PiaNote does improve user's sight-reading confidence and abilities, but further research must be conducted to clearly validate PiaNote's effectiveness. We conclude that PiaNote has potential to become an effective sight-reading application with slight improvements and further research.</p>

Name thesis:degree_name

MS in Computer Science

Discipline thesis:degree_discipline

Computer Science

Year dc:date.available

2016

Author dc:creator

• Schulz, Drew

Contributors dc:contributor

• John Clements

• music
• piano
• algorithm
• edit-distance
• sight-reading
• AI
• Music Education

Identifier

10.15368/theses.2016.53

OAI identifier oai:identifier

oai:digitalcommons.calpoly.edu:theses-2765