Sheet Music Digitization

This master thesis presents a novel approach to improving Optical Music Recognition (OMR) systems for sheet music digitization. The proposed solution involves incorporating a new module to address ambiguous decisions in music symbol recognition, enhancing the accuracy and efficiency of the transcription process. Through a comparative analysis with existing OMR systems and performance testing using a representative dataset, the thesis seeks to demonstrate the effectiveness of the proposed solution in overcoming challenges related to complex music notations and improving recognition capabilities. By focusing on addressing ambiguous decisions in OMR, this research contributes to advancing the field of sheet music digitization and lays the foundation for more robust and accurate OMR systems in the future.