EDMSound: Spectrogram Based Diffusion Models for Efficient and High-Quality Audio Synthesis
Abstract
Audio diffusion models can synthesize a wide variety of sounds. Existing models often operate on the latent domain with cascaded phase recovery modules to reconstruct waveform. This poses challenges when generating high-fidelity audio. In this paper, we propose EDMSound, a diffusion-based generative model in spectrogram domain under the framework of elucidated diffusion models (EDM). Combining with efficient deterministic sampler, we achieved similar Fr\'echet audio distance (FAD) score as top-ranked baseline with only 10 steps and reached state-of-the-art performance with 50 steps on the DCASE2023 foley sound generation benchmark. We also revealed a potential concern regarding diffusion based audio generation models that they tend to generate samples with high perceptual similarity to the data from training data. Project page: https://agentcooper2002.github.io/EDMSound/
Cite
@article{arxiv.2311.08667,
title = {EDMSound: Spectrogram Based Diffusion Models for Efficient and High-Quality Audio Synthesis},
author = {Ge Zhu and Yutong Wen and Marc-André Carbonneau and Zhiyao Duan},
journal= {arXiv preprint arXiv:2311.08667},
year = {2023}
}
Comments
Accepted at NeurIPS Workshop: Machine Learning for Audio (Camera Ready)