Saved in:
Bibliographic Details
Main Authors: Zhou, Feiyan, Wang, Luyuan, Chen, Shoufa, Wang, Zhe, Liu, Zhiheng, Cong, Yuren, Zhang, Xiaohui, Yang, Fanny, Zeng, Belinda
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2605.18749
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917508477353984
author Zhou, Feiyan
Wang, Luyuan
Chen, Shoufa
Wang, Zhe
Liu, Zhiheng
Cong, Yuren
Zhang, Xiaohui
Yang, Fanny
Zeng, Belinda
author_facet Zhou, Feiyan
Wang, Luyuan
Chen, Shoufa
Wang, Zhe
Liu, Zhiheng
Cong, Yuren
Zhang, Xiaohui
Yang, Fanny
Zeng, Belinda
contents Modern audio generation predominantly relies on latent-space compression, introducing additional complexity and potential information loss. In this work, we challenge this paradigm with WavFlow, a framework that generates high-fidelity audio directly in raw waveform space without intermediate representations. To overcome the inherent difficulties of modeling high-dimensional and low-energy signals, we reshape audio into 2D token grids through waveform patchify and introduce amplitude lifting to align signal scales, enabling stable optimization via direct x-prediction in flow matching. To capture complex semantic alignment and temporal synchronization, we leverage an automated data pipeline to curate 5 million high-quality video-text-audio triplets, allowing the model to learn fine-grained acoustic patterns from scratch. Experimental results show that WavFlow achieves competitive performance on the video-to-audio benchmark VGGSound (FD_PaSST: 59.98, IS_PANNs: 17.40, DeSync: 0.44) and the text-to-audio benchmark AudioCaps (FD_PANNs: 10.63, IS_PANNs: 12.62), matching or exceeding the performance of established latent-based methods. Our work demonstrates that intermediate compression is not a prerequisite for high-quality synthesis, offering a simpler and more scalable alternative for multimodal audio generation.
format Preprint
id arxiv_https___arxiv_org_abs_2605_18749
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle WavFlow: Audio Generation in Waveform Space
Zhou, Feiyan
Wang, Luyuan
Chen, Shoufa
Wang, Zhe
Liu, Zhiheng
Cong, Yuren
Zhang, Xiaohui
Yang, Fanny
Zeng, Belinda
Sound
Computer Vision and Pattern Recognition
Modern audio generation predominantly relies on latent-space compression, introducing additional complexity and potential information loss. In this work, we challenge this paradigm with WavFlow, a framework that generates high-fidelity audio directly in raw waveform space without intermediate representations. To overcome the inherent difficulties of modeling high-dimensional and low-energy signals, we reshape audio into 2D token grids through waveform patchify and introduce amplitude lifting to align signal scales, enabling stable optimization via direct x-prediction in flow matching. To capture complex semantic alignment and temporal synchronization, we leverage an automated data pipeline to curate 5 million high-quality video-text-audio triplets, allowing the model to learn fine-grained acoustic patterns from scratch. Experimental results show that WavFlow achieves competitive performance on the video-to-audio benchmark VGGSound (FD_PaSST: 59.98, IS_PANNs: 17.40, DeSync: 0.44) and the text-to-audio benchmark AudioCaps (FD_PANNs: 10.63, IS_PANNs: 12.62), matching or exceeding the performance of established latent-based methods. Our work demonstrates that intermediate compression is not a prerequisite for high-quality synthesis, offering a simpler and more scalable alternative for multimodal audio generation.
title WavFlow: Audio Generation in Waveform Space
topic Sound
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2605.18749