# A Study of the Scale Invariant Signal to Distortion Ratio in Speech Separation with Noisy References

arXiv: https://arxiv.org/abs/2508.14623 · HTML: https://mlsanity.com/paper/2508.14623/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 37 of its 59 references that are in the map.
- First posted on arXiv 20 Aug 2025.

## Builds on (most cited first, 30 of 37)

- [SDR - half-baked or well done?](https://mlsanity.com/paper/1811.02508/) (2018)
- [DNSMOS: A Non-Intrusive Perceptual Objective Speech Quality metric to evaluate Noise Suppressors](https://mlsanity.com/paper/2010.15258/) (2020)
- [Supervised Speech Separation Based on Deep Learning: An Overview](https://mlsanity.com/paper/1708.07524/) (2017)
- [Deep clustering: Discriminative embeddings for segmentation and separation](https://mlsanity.com/paper/1508.04306/) (2015)
- [NISQA: A Deep CNN-Self-Attention Model for Multidimensional Speech Quality Prediction with Crowdsourced Datasets](https://mlsanity.com/paper/2104.09494/) (2021)
- [Permutation Invariant Training of Deep Models for Speaker-Independent Multi-talker Speech Separation](https://mlsanity.com/paper/1607.00325/) (2016)
- [Attention is All You Need in Speech Separation](https://mlsanity.com/paper/2010.13154/) (2020)
- [Dual-path RNN: efficient long sequence modeling for time-domain single-channel speech separation](https://mlsanity.com/paper/1910.06379/) (2019)
- [WHAM!: Extending Speech Separation to Noisy Environments](https://mlsanity.com/paper/1907.01160/) (2019)
- [LibriMix: An Open-Source Dataset for Generalizable Speech Separation](https://mlsanity.com/paper/2005.11262/) (2020)
- [TasNet: time-domain audio separation network for real-time, single-channel speech separation](https://mlsanity.com/paper/1711.00541/) (2017)
- [MetricGAN+: An Improved Version of MetricGAN for Speech Enhancement](https://mlsanity.com/paper/2104.03538/) (2021)
- [Wavesplit: End-to-End Speech Separation by Speaker Clustering](https://mlsanity.com/paper/2002.08933/) (2020)
- [Dual-Path Transformer Network: Direct Context-Aware Modeling for End-to-End Monaural Speech Separation](https://mlsanity.com/paper/2007.13975/) (2020)
- [EARS: An Anechoic Fullband Speech Dataset Benchmarked for Speech Enhancement and Dereverberation](https://mlsanity.com/paper/2406.06185/) (2024)
- [WHAMR!: Noisy and Reverberant Single-Channel Speech Separation](https://mlsanity.com/paper/1910.10279/) (2019)
- [Open-Source Conversational AI with SpeechBrain 1.0](https://mlsanity.com/paper/2407.00463/) (2024)
- [Sudo rm -rf: Efficient Networks for Universal Audio Source Separation](https://mlsanity.com/paper/2007.06833/) (2020)
- [TF-GridNet: Making Time-Frequency Domain Models Great Again for Monaural Speaker Separation](https://mlsanity.com/paper/2209.03952/) (2022)
- [Single-Channel Multi-Speaker Separation using Deep Clustering](https://mlsanity.com/paper/1607.02173/) (2016)
- [MossFormer: Pushing the Performance Limit of Monaural Speech Separation using Gated Single-Head Transformer with Convolution-Augmented Joint Self-Attentions](https://mlsanity.com/paper/2302.11824/) (2023)
- [Divide and Conquer: A Deep CASA Approach to Talker-independent Monaural Speaker Separation](https://mlsanity.com/paper/1904.11148/) (2019)
- [Speaker-independent Speech Separation with Deep Attractor Network](https://mlsanity.com/paper/1707.03634/) (2017)
- [FurcaNeXt: End-to-end monaural speech separation with dynamic gated dilated temporal convolutional networks](https://mlsanity.com/paper/1902.04891/) (2019)
- [An empirical study of Conv-TasNet](https://mlsanity.com/paper/2002.08688/) (2020)
- [On Cross-Corpus Generalization of Deep Learning Based Speech Enhancement](https://mlsanity.com/paper/2002.04027/) (2020)
- [TF-Locoformer: Transformer with Local Modeling by Convolution for Speech Separation and Enhancement](https://mlsanity.com/paper/2408.03440/) (2024)
- [End-to-End Speech Separation with Unfolded Iterative Phase Reconstruction](https://mlsanity.com/paper/1804.10204/) (2018)
- [Analysis of Deep Clustering as Preprocessing for Automatic Speech Recognition of Sparsely Overlapping Speech](https://mlsanity.com/paper/1905.03500/) (2019)
- [Mixup-breakdown: a consistency training method for improving generalization of speech separation models](https://mlsanity.com/paper/1910.13253/) (2019)

## Built on by (most cited first, 4 of 4)

- [Exploring Resolution-Wise Shared Attention in Hybrid Mamba-U-Nets for Improved Cross-Corpus Speech Enhancement](https://mlsanity.com/paper/2510.01958/) (2025)
- [DynFOA: Generating First-Order Ambisonics with Conditional Diffusion for Dynamic and Acoustically Complex 360-Degree Videos](https://mlsanity.com/paper/2602.06846/) (2026)
- [VorTEX: Various overlap ratio for Target speech EXtraction](https://mlsanity.com/paper/2603.14803/) (2026)
- [DynFOA: Generating First-Order Ambisonics with Conditional Diffusion for Dynamic and Acoustically Complex 360-Degree Videos](https://mlsanity.com/paper/2604.02781/) (2026)

## Abstract (arXiv)

This paper examines the implications of using the Scale-Invariant Signal-to-Distortion Ratio (SI-SDR) as both evaluation and training objective in supervised speech separation, when the training references contain noise, as is the case with the de facto benchmark WSJ0-2Mix. A derivation of the SI-SDR with noisy references reveals that noise limits the achievable SI-SDR, or leads to undesired noise in the separated outputs. To address this, a method is proposed to enhance references and augment the mixtures with WHAM!, aiming to train models that avoid learning noisy references. Two models trained on these enhanced datasets are evaluated with the non-intrusive NISQA.v2 metric. Results show reduced noise in separated speech but suggest that processing references may introduce artefacts, limiting overall quality gains. Negative correlation is found between SI-SDR and perceived noisiness across models on the WSJ0-2Mix and Libri2Mix test sets, underlining the conclusion from the derivation.