RenderBox is a text-and-score conditioned diffusion model that renders expressive, controllable audio performances across piano, guitar, saxophone, violin, and orchestral instruments.
From MNIST to ImageNet and back: benchmarking continual curriculum learning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.AS 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
RenderBox: Expressive Performance Rendering with Text Control
RenderBox is a text-and-score conditioned diffusion model that renders expressive, controllable audio performances across piano, guitar, saxophone, violin, and orchestral instruments.