音乐抄袭检测数据集(MPD-Set)
收藏资源简介:
音乐抄袭检测数据集(MPD-Set)是由上海交通大学教育部人工智能重点实验室的研究人员在知名国家级音乐领域专业机构的指导下创建的,旨在解决音乐抄袭检测问题。该数据集包含2000首音乐作品,每对作品之间存在抄袭关系,涵盖了旋律、节奏和音调层面的多种实际抄袭案例。数据集的创建过程涉及从Wikifonia开放源数据集中提取音乐片段,并将其转换为MIDI格式。MPD-Set的应用领域主要集中在音乐版权保护和抄袭检测,通过提供详细的数据支持,帮助识别和量化音乐作品中的抄袭行为。
The Music Plagiarism Detection Dataset (MPD-Set) was developed by researchers from the Key Laboratory of Artificial Intelligence of the Ministry of Education, Shanghai Jiao Tong University, under the guidance of well-known national-level professional music institutions, aiming to address the problem of music plagiarism detection. This dataset includes 2000 musical works, with each pair of works having a plagiarism relationship, covering various actual plagiarism cases at the levels of melody, rhythm and pitch. The dataset creation process involves extracting musical segments from the Wikifonia open-source dataset and converting them into MIDI format. The main application fields of MPD-Set focus on music copyright protection and plagiarism detection, providing detailed data support to help identify and quantify plagiarism in musical works.




