arXiv：cs.AI· Bowen Zhang, Junchuan Zhao, Ian McLoughlin, Ye Wang, A S Madhukumar·· 5 小时前
CodecFlow：在神经编解码器潜在空间用条件流匹配实现高效带宽扩展
CodecFlow: Efficient Bandwidth Extension via Conditional Flow Matching in Neural Codec Latent Space
AI 导读
CodecFlow 是一个基于神经音频编解码器的语音带宽扩展（BWE）框架，在紧凑潜在空间中完成高效语音重建。它采用音高感知的条件流转换器和结构约束残差向量量化器来提升潜在对齐稳定性，端到端优化后在 8 kHz 到 16 kHz 及 44.1 kHz 语音 BWE 任务上取得较强的频谱保真度与感知质量。
正文
Abstract:Speech Bandwidth Extension improves clarity and intelligibility by restoring/inferring appropriate high-frequency content for low-bandwidth speech. Existing methods often rely on spectrogram or waveform modeling, which can incur higher computational cost and have limited high-frequency fidelity. Neural audio codecs offer compact latent representations that better preserve acoustic detail, yet accurately recovering high-resolution latent information remains challenging due to representation mismatch. We present CodecFlow, a neural codec-based BWE framework that performs efficient speech reconstruction in a compact latent space. CodecFlow employs a voicing-aware conditional flow converter on continuous codec embeddings and a structure-constrained residual vector quantizer to improve latent alignment stability. Optimized end-to-end, CodecFlow achieves strong spectral fidelity and enhanced perceptual quality on 8 kHz to 16 kHz and 44.1 kHz speech BWE tasks.
| Comments: | 7 pages, 7 figures |
| Subjects: | Sound (cs.SD); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2603.02022 [cs.SD] |
| (or arXiv:2603.02022v3 [cs.SD] for this version) | |
| https://doi.org/10.48550/arXiv.2603.02022 arXiv-issued DOI via DataCite |
Submission history
From: Bowen Zhang [view email]
[v1]
Mon, 2 Mar 2026 16:03:46 UTC (4,563 KB)
[v2]
Tue, 3 Mar 2026 06:49:12 UTC (4,563 KB)
[v3]
Sun, 11 Oct 2026 14:51:42 UTC (4,564 KB)
来源：arXiv：cs.AI · arxiv.org