InfiniteAudio: Infinite-Length Audio Generation with Consistency
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915321744457728 |
|---|---|
| author | Jung, Chaeyoung Ki, Hojoon Kim, Ji-Hoon Kim, Junmo Chung, Joon Son |
| author_facet | Jung, Chaeyoung Ki, Hojoon Kim, Ji-Hoon Kim, Junmo Chung, Joon Son |
| contents | This paper presents InfiniteAudio, a simple yet effective strategy for generating infinite-length audio using diffusion-based text-to-audio methods. Current approaches face memory constraints because the output size increases with input length, making long duration generation challenging. A common workaround is to concatenate short audio segments, but this often leads to inconsistencies due to the lack of shared temporal context. To address this, InfiniteAudio integrates seamlessly into existing pipelines without additional training. It introduces two key techniques: FIFO sampling, a first-in, first-out inference strategy with fixed-size inputs, and curved denoising, which selectively prioritizes key diffusion steps for efficiency. Experiments show that InfiniteAudio achieves comparable or superior performance across all metrics. Audio samples are available on our project page. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_03020 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | InfiniteAudio: Infinite-Length Audio Generation with Consistency Jung, Chaeyoung Ki, Hojoon Kim, Ji-Hoon Kim, Junmo Chung, Joon Son Audio and Speech Processing This paper presents InfiniteAudio, a simple yet effective strategy for generating infinite-length audio using diffusion-based text-to-audio methods. Current approaches face memory constraints because the output size increases with input length, making long duration generation challenging. A common workaround is to concatenate short audio segments, but this often leads to inconsistencies due to the lack of shared temporal context. To address this, InfiniteAudio integrates seamlessly into existing pipelines without additional training. It introduces two key techniques: FIFO sampling, a first-in, first-out inference strategy with fixed-size inputs, and curved denoising, which selectively prioritizes key diffusion steps for efficiency. Experiments show that InfiniteAudio achieves comparable or superior performance across all metrics. Audio samples are available on our project page. |
| title | InfiniteAudio: Infinite-Length Audio Generation with Consistency |
| topic | Audio and Speech Processing |
| url | https://arxiv.org/abs/2506.03020 |