Even with strong sequence models like Transformers, generating expressive piano performances with long-range musical structures remains challenging. Meanwhile, methods to compose well-structured melodies or lead sheets (melody + chords), i.e., simpler forms of music, gained more success. Observing the above, we devise a two-stage Transformer-based framework that Composes a lead sheet first, and then Embellishes it with accompaniment and expressive touches. Such a factorization also enables pretraining on non-piano data. Our objective and subjective experiments show that Compose & Embellish shrinks the gap in structureness between a current state of the art and real performances by half, and improves other musical aspects such as richness and coherence as well.

翻译：尽管有诸如Transformer等强大的序列模型，生成具有长程音乐结构的富有表现力的钢琴表演仍然具有挑战性。同时，为作曲良好结构的旋律或导谱（旋律+和弦）——即更简单的音乐形式——的方法取得了更多成功。基于上述观察，我们设计了一个两阶段基于Transformer的框架，首先作曲导谱，然后通过伴奏和表现性细节对其进行装饰。这种分解还使得能够对非钢琴数据进行预训练。我们的客观和主观实验表明，创作与装饰将当前最先进模型与真实表演之间的结构性差距缩小了一半，并改善了音乐的丰富性和连贯性等其他方面。