# Efficient And Scalable Neural Residual Waveform Coding With Collaborative Quantization

arXiv: https://arxiv.org/abs/2002.05604 · HTML: https://mlsanity.com/paper/2002.05604/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 13 Feb 2020.

## Built on by (most cited first, 7 of 7)

- [HARP-Net: Hyper-Autoencoded Reconstruction Propagation for Scalable Neural Audio Coding](https://mlsanity.com/paper/2107.10843/) (2021)
- [Scalable and Efficient Neural Speech Coding: A Hybrid Design](https://mlsanity.com/paper/2103.14776/) (2021)
- [End-to-End Binaural Speech Synthesis](https://mlsanity.com/paper/2207.03697/) (2022)
- [Sub-8-bit quantization for on-device speech recognition: a regularization-free approach](https://mlsanity.com/paper/2210.09188/) (2022)
- [Cross-Scale Vector Quantization for Scalable Neural Speech Coding](https://mlsanity.com/paper/2207.03067/) (2022)
- [NESC: Robust Neural End-2-End Speech Coding with GANs](https://mlsanity.com/paper/2207.03282/) (2022)
- [FlowMAC: Conditional Flow Matching for Audio Coding at Low Bit Rates](https://mlsanity.com/paper/2409.17635/) (2024)

## Abstract (arXiv)

Scalability and efficiency are desired in neural speech codecs, which supports a wide range of bitrates for applications on various devices. We propose a collaborative quantization (CQ) scheme to jointly learn the codebook of LPC coefficients and the corresponding residuals. CQ does not simply shoehorn LPC to a neural network, but bridges the computational capacity of advanced neural network models and traditional, yet efficient and domain-specific digital signal processing methods in an integrated manner. We demonstrate that CQ achieves much higher quality than its predecessor at 9 kbps with even lower model complexity. We also show that CQ can scale up to 24 kbps where it outperforms AMR-WB and Opus. As a neural waveform codec, CQ models are with less than 1 million parameters, significantly less than many other generative models.