Authors:

Edward Falcon Navarro

Date Created:

Not specified

Course Title:

Professor:

Not specified

About Paper:

Musicians practice together or alone to refine their skills; however with various scheduling conflicts and geographical distance issues, group practice can be a struggle. The development of Companion has been aimed to address these conflicts. One general problem with various other accompaniment systems is interaction with the device running the system. On the fly, changes are currently quite difficult especially considering one has to operate an instrument while performing. With that said, the development of an efficient voice commanding system has resolved this issue, enabling users to interact with the application hands-free. In the pursuit of the best possible vocal recognition library, we explored various offline and online libraries on accuracy and runtime. Our results determined the library we chose had the best balance of speed as well as accuracy for musicians to practice with. Future work could expand this robust system of user defined voice commands to various other teams inside of Artificial Intelligence for Musicians. Keywords: Musical Accompaniment; Large Language Models; Voice Recognition; Audio Synthesis

Source:

Purdue University / 2024

Topics:

No topics listed

Co-authors:

Edward Falcon Navarro