GAPS (Guitar-Aligned Performance Scores) Dataset

Release of the aligned MIDI transcriptions, scores and downbeats that constitute the GAPS dataset. Links to YouTube URLs for audio and video are provided in the accompanying metadata file. See below for the abstract of the publication. Abstract: We introduce GAPS (Guitar-Aligned Performance Scores), a new dataset of classical guitar performances, and a benchmark guitar transcription model that achieves state-of-the-art performance on GuitarSet in both supervised and zero-shot settings. GAPS is the largest dataset of real guitar audio, containing 14 hours of freely available audio-score aligned pairs, recorded in diverse conditions by over 200 performers, together with high-resolution note-level MIDI alignments and performance videos. These enable us to train a state-of-the-art model for automatic transcription of solo guitar recordings which can generalise well to real world audio that is unseen during training.本数据集发布了构成GAPS (Guitar-Aligned Performance Scores) 数据集的对齐MIDI转录谱、乐谱与强拍（downbeats）标注。配套元数据文件中提供了音频与视频对应的YouTube链接。下文将展示该出版物的摘要。 摘要： 我们推出了GAPS数据集与一款基准吉他转录模型：该数据集为全新古典吉他演奏数据集，而该模型在监督学习与零样本（zero-shot）设置下的GuitarSet任务上均实现了当前最优性能。GAPS是目前规模最大的真实吉他音频数据集，包含14小时可免费获取的音频-乐谱对齐样本，由200余名演奏者在多样化的录制环境中采集，同时附带高分辨率音符级MIDI对齐信息与演奏视频。这些资源可支持我们训练出一款当前最优的独奏吉他录音自动转录模型，该模型能够良好泛化至训练阶段未见过的真实世界音频。