Music102: An $D_{12}$-equivariant transformer for chord progression accompaniment
This work addresses challenges in computational music analysis for music generation and composition tools, though it is incremental as it builds on prior methods with specific enhancements.
The paper tackled the problem of chord progression accompaniment by developing Music102, a D12-equivariant transformer that leverages musical symmetry, resulting in significant improvements over a non-equivariant prototype in weighted loss and exact accuracy metrics on the POP909 dataset while using fewer parameters.
We present Music102, an advanced model aimed at enhancing chord progression accompaniment through a $D_{12}$-equivariant transformer. Inspired by group theory and symbolic music structures, Music102 leverages musical symmetry--such as transposition and reflection operations--integrating these properties into the transformer architecture. By encoding prior music knowledge, the model maintains equivariance across both melody and chord sequences. The POP909 dataset was employed to train and evaluate Music102, revealing significant improvements over the non-equivariant Music101 prototype Music101 in both weighted loss and exact accuracy metrics, despite using fewer parameters. This work showcases the adaptability of self-attention mechanisms and layer normalization to the discrete musical domain, addressing challenges in computational music analysis. With its stable and flexible neural framework, Music102 sets the stage for further exploration in equivariant music generation and computational composition tools, bridging mathematical theory with practical music performance.