Gangqinpu dataset enables diverse multi-modal piano music analysis
Gangqinpu: A Multi-Genre, Multi-Task Piano Dataset with Aligned Audio, MIDI, and Musical Scores
SoundArtificial Intelligence
Summary
Understanding how piano music sounds, is played, and is written is hard because datasets that connect recorded sound, piano performance data, and written music are rare and small. The authors created Gangqinpu, a large collection of piano music recordings from many styles like pop, jazz, and classical, matched with musical notes and performance details. They used a new method to link these different kinds of data and showed that Gangqinpu helps improve tasks like turning performances into sheet music and recognizing musical beats. Their work provides a resource for building better music software and tools.
What this means in practice
- •For music software developers: Train and benchmark piano transcription and beat tracking algorithms using diverse aligned audio, MIDI, and score data across multiple genres.
- •For digital music educators: Create tools that assess and display expressive piano performances by comparing recordings to written music with precise alignment.
Authors
Wei Zeng, Anvay Mathur, Ye Wang
Abstract
Aligned audio, performance MIDI, and musical scores provide valuable supervision for music analysis and performance modeling, yet datasets that jointly align all three modalities remain scarce and limited in scale and genre diversity. We introduce Gangqinpu, a multi-genre piano dataset containing 18,786 score-performance pairs and 1,007.1 hours of source audio, with automatic note-level alignments and derived beat/downbeat annotations. It spans diverse genres, including pop, classical, jazz, folk/traditional, rock, and blues. Our automated pipeline transcribes piano audio, aligns the resulting MIDI with musical scores, assigns genre labels, and filters out non-expressive performances with a score-informed binary classifier. We further construct joint splits with ASAP using conservative score-similarity grouping to reduce data leakage. We evaluate Gangqinpu on performance MIDI-to-score transcription, expressive performance rendering, and beat tracking. Downstream experiments demonstrate the value of Gangqinpu as complementary training data, with cross-dataset gains in beat tracking and primarily in-domain benefits for transcription and expressive performance rendering. The code and dataset are publicly available at https://github.com/wei-zeng98/gangqinpu.