arXiv cs.AI· Zijian Lin, Jing Yang, Jinghao Luo, Zhuo Wang, Fan Fan, Zhiyong Wu·· 3 天前AI 评分22
BridgeCodec：Mamba 增强的 Schrödinger Bridge 神经音频编解码器，实现低比特率跨编解码器转换
BridgeCodec: Mamba Enhanced Neural Audio Codec with Schr\"odinger Bridge at Low Bitrate
AI 导读
BridgeCodec 通过将跨编解码器转换建模为 Schrödinger Bridge 最优传输问题，解耦了冻结的编码器-解码器对，并集成 Mamba 增强的 U-Net 架构捕捉长程语音时序依赖。在 1 kbps 超低比特率下，该框架将 8 kHz 轻量源编码器直接映射到 48 kHz 目标解码器，实现了优异的宽带重建和感知质量。该工作已被 Interspeech 2026 接收。
正文 · 原文
Abstract:Neural audio codecs achieve remarkable performance but operate as closed ecosystems with rigidly co-trained encoderdecoder pairs, severely hindering interoperability. To resolve this, we propose BridgeCodec, a versatile framework that decouples these pairs by translating latent representations between mismatched, frozen endpoints. We formulate this cross-codec translation as a Schrödinger Bridge optimal transport problem, enabling robust mapping between disparate distributions without predefined noise priors. To efficiently capture longrange speech temporal dependencies, we integrate a Mambaenhanced U-Net architecture. We validate BridgeCodec on an extreme deployment scenario: mapping a lightweight 8 kHz source encoder directly to a 48 kHz target decoder at an ultralow bitrate of 1 kbps. Experimental results demonstrate that BridgeCodec achieves superior wideband reconstruction and high perceptual quality, successfully bridging heavily bottlenecked, heterogeneous codecs.
| Comments: | Accepted by Interspeech 2026 |
| Subjects: | Sound (cs.SD); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Audio and Speech Processing (eess.AS) |
| Cite as: | arXiv:2610.15453 [cs.SD] |
| (or arXiv:2610.15453v1 [cs.SD] for this version) | |
| https://doi.org/10.48550/arXiv.2610.15453 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Jing Yang [view email]
[v1]
Sun, 11 Oct 2026 08:54:53 UTC (151 KB)
来源：arXiv cs.AI · arxiv.org