Semantic Scholar Open Access 2024 213 sitasi

Fast Timing-Conditioned Latent Audio Diffusion

Zach Evans CJ Carr Josiah Taylor Scott H. Hawley Jordi Pons

Abstrak

Generating long-form 44.1kHz stereo audio from text prompts can be computationally demanding. Further, most previous works do not tackle that music and sound effects naturally vary in their duration. Our research focuses on the efficient generation of long-form, variable-length stereo music and sounds at 44.1kHz using text prompts with a generative model. Stable Audio is based on latent diffusion, with its latent defined by a fully-convolutional variational autoencoder. It is conditioned on text prompts as well as timing embeddings, allowing for fine control over both the content and length of the generated music and sounds. Stable Audio is capable of rendering stereo signals of up to 95 sec at 44.1kHz in 8 sec on an A100 GPU. Despite its compute efficiency and fast inference, it is one of the best in two public text-to-music and -audio benchmarks and, differently from state-of-the-art models, can generate music with structure and stereo sounds.

Topik & Kata Kunci

Computer Science Engineering

Penulis (5)

Zach Evans

CJ Carr

Josiah Taylor

Scott H. Hawley

Jordi Pons

Format Sitasi

APA MLA BibTeX

Evans, Z., Carr, C., Taylor, J., Hawley, S.H., Pons, J. (2024). Fast Timing-Conditioned Latent Audio Diffusion. https://doi.org/10.48550/arXiv.2402.04825

Akses Cepat

PDF tidak tersedia langsung

Cek di sumber asli →

Lihat di Sumber doi.org/10.48550/arXiv.2402.04825

Informasi Jurnal

Tahun Terbit: 2024
Bahasa: en
Total Sitasi: 213×
Sumber Database: Semantic Scholar
DOI: 10.48550/arXiv.2402.04825
Akses: Open Access ✓