MIDI music plagiarism detection method based on feature similarity learning

Because of increasing plagiarism of Musical Instrument Digital Interface (MIDI) music, we proposed a MIDI music plagiarism detection method through music feature similarity learning. In order to extract more representative key features of music, further research is needed on music feature extraction and enhancement methods. This can improve the performance of music similarity detection. The proposed method extracts rhythm and note vectors from MIDI music as features and detects similarity using a Siamese network. Experimental results show that the proposed method has a good effect on the similarity detection of MIDI music, and the accuracy reaches 96.3%. This method effectively detects the similarity between different MIDI music, which provides reliable technology support for evaluating the plagiarism between musical works.