# Speech Separation using Neural Audio Codecs with Embedding Loss

arXiv: https://arxiv.org/abs/2411.17998 · HTML: https://mlsanity.com/paper/2411.17998/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 27 Nov 2024.

## Built on by (most cited first, 3 of 3)

- [Learning to Upsample and Upmix Audio in the Latent Domain](https://mlsanity.com/paper/2506.00681/) (2025)
- [Re-Bottleneck: Latent Re-Structuring for Neural Audio Autoencoders](https://mlsanity.com/paper/2507.07867/) (2025)
- [Unsupervised Single-Channel Speech Separation with Diffusion under Speaker-Embedding Guidance](https://mlsanity.com/paper/2509.24395/) (2025)

## Abstract (arXiv)

Neural audio codecs have revolutionized audio processing by enabling speech tasks to be performed on highly compressed representations. Recent work has shown that speech separation can be achieved within these compressed domains, offering faster training and reduced inference costs. However, current approaches still rely on waveform-based loss functions, necessitating unnecessary decoding steps during training. We propose a novel embedding loss for neural audio codec-based speech separation that operates directly on compressed audio representations, eliminating the need for decoding during training. To validate our approach, we conduct comprehensive evaluations using both objective metrics and perceptual assessment techniques, including intrusive and non-intrusive methods. Our results demonstrate that embedding loss can be used to train codec-based speech separation models with a 2x improvement in training speed and computational cost while achieving better DNSMOS and STOI performance on the WSJ0-2mix dataset across 3 different pre-trained codecs.