MuSViT: A Foundation Vision Model for Sheet Music Representation

MuSViT: A Foundation Vision Model for Sheet Music Representation

MuSViT is a vision transformer-based foundation model pre-trained on millions of sheet music pages that demonstrates superior performance in music score recognition and symbol dete…

Hugging Face · Daily Papers ·Carlos Penarrubia, Antonio Rios-Vila · ·▲ 2 upvotes

Este artigo está em destaque na seleção diária de papers do Hugging Face, curada pela comunidade de pesquisa em IA.

Autores: Carlos Penarrubia, Antonio Rios-Vila, Eliseo Fuentes-Martinez, Juan C. Martinez-Sevilla, Francisco J. Castellanos, María Alfaro-Contreras

  • 2 upvotes da comunidade
  • Temas: vision transformer, foundation model, Masked Autoencoders, ViT encoder, sheet music, music score recognition

Resumo

Resumo original (em inglês), extraído do paper:

MuSViT is a vision transformer-based foundation model pre-trained on millions of sheet music pages that demonstrates superior performance in music score recognition and symbol detection tasks through both linear probing and fine-tuning approaches.

Onde ler

compartilhar: