多尺度量化提升音频压缩效率，适配不同时间尺度的音频结构。
SNAC: Multi-Scale Neural Audio Codec
- 在不同帧率下使用分层量化器，捕捉多尺度音频特征。
- 相比传统方法，在低码率下实现更高音质，主观评测得分提升显著。
- 适合语音生成、音频压缩等需要高效表示的场景。
神经音频编解码器因其能在极低码率下实现高保真音频表示而受到关注，使得语言建模方法可用于音频生成与理解。残差向量量化（RVQ）已成为神经音频压缩的标准技术，通过级联多个向量量化代码本实现。本文提出多尺度神经音频编解码器（SNAC），是对RVQ的简单扩展，允许量化器在不同时间分辨率下运行。通过在可变帧率下应用分层量化器，编解码器能够适应音频在多时间尺度上的结构变化。大量客观与主观评估表明，该方法实现了更高效的压缩性能。代码与模型权重已开源至 https://github.com/hubertsiuzdak/snac。
原文摘要 · Abstract (English)
Neural audio codecs have recently gained popularity because they can represent audio signals with high fidelity at very low bitrates, making it feasible to use language modeling approaches for audio generation and understanding. Residual Vector Quantization (RVQ) has become the standard technique for neural audio compression using a cascade of VQ codebooks. This paper proposes the Multi-Scale Neural Audio Codec, a simple extension of RVQ where the quantizers can operate at different temporal resolutions. By applying a hierarchy of quantizers at variable frame rates, the codec adapts to the audio structure across multiple timescales. This leads to more efficient compression, as demonstrated by extensive objective and subjective evaluations. The code and model weights are open-sourced at https://github.com/hubertsiuzdak/snac.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。