From Discrete T okens to High-Fidelity Audio Using Multi-Band Diffusion Robin San Roman

Oct-8-2025, 00:34:37 GMT–Neural Information Processing Systems

Deep generative models can generate high-fidelity audio conditioned on various types of representations (e.g., mel-spectrograms, Mel-frequency Cepstral Coefficients (

arxiv preprint arxiv, machine learning, natural language, (17 more...)

Neural Information Processing Systems

Oct-8-2025, 00:34:37 GMT

Conferences PDF

Country:
- Asia > Middle East
  - Israel > Jerusalem District > Jerusalem (0.04)
- Europe
  - France > Grand Est
    - Meurthe-et-Moselle > Nancy (0.04)
  - Italy > Calabria
    - Catanzaro Province > Catanzaro (0.04)
- South America > Chile
  - Santiago Metropolitan Region > Santiago Province > Santiago (0.04)

Genre:
- Research Report (0.68)

Industry:
- Leisure & Entertainment (0.47)

Technology:
- Information Technology > Artificial Intelligence
  - Machine Learning > Neural Networks
    - Deep Learning > Generative AI (0.34)
  - Natural Language (1.00)
  - Speech (1.00)

Duplicate Docs Excel Report

Title
FromDiscreteTokenstoHigh-FidelityAudioUsing Multi-BandDiffusion

Similar Docs Excel Report more

Title	Similarity	Source
None found