# Data augmentation and loss normalization for deep noise suppression

arXiv: https://arxiv.org/abs/2008.06412 · HTML: https://mlsanity.com/paper/2008.06412/

## Key facts

- Cited by 23 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 14 Aug 2020.

## Built on by (most cited first, 23 of 23)

- [SpeechT5: Unified-Modal Encoder-Decoder Pre-Training for Spoken Language Processing](https://mlsanity.com/paper/2110.07205/) (2021)
- [MaskSR: Masked Language Model for Full-band Speech Restoration](https://mlsanity.com/paper/2406.02092/) (2024)
- [AnyEnhance: A Unified Generative Model with Prompt-Guidance and Self-Critic for Voice Enhancement](https://mlsanity.com/paper/2501.15417/) (2025)
- [Real-time Denoising and Dereverberation with Tiny Recurrent U-Net](https://mlsanity.com/paper/2102.03207/) (2021)
- [A Generalized Surface Loss for Reducing the Hausdorff Distance in Medical Imaging Segmentation](https://mlsanity.com/paper/2302.03868/) (2023)
- [Dynamic nsNet2: Efficient Deep Noise Suppression with Early Exiting](https://mlsanity.com/paper/2308.16678/) (2023)
- [Ultra Low Complexity Deep Learning Based Noise Suppression](https://mlsanity.com/paper/2312.08132/) (2023)
- [Complex-Valued Time-Frequency Self-Attention for Speech Dereverberation](https://mlsanity.com/paper/2211.12632/) (2022)
- [A Variance-Preserving Interpolation Approach for Diffusion Models with Applications to Single Channel Speech Enhancement and Recognition](https://mlsanity.com/paper/2405.16952/) (2024)
- [Adaptive Slimming for Scalable and Efficient Speech Enhancement](https://mlsanity.com/paper/2507.04879/) (2025)
- [CheapNET: Improving Light-weight speech enhancement network by projected loss function](https://mlsanity.com/paper/2311.15959/) (2023)
- [Hold Me Tight: Stable Encoder-Decoder Design for Speech Enhancement](https://mlsanity.com/paper/2408.17358/) (2024)
- [MoE-Gyro: Self-Supervised Over-Range Reconstruction and Denoising for MEMS Gyroscopes](https://mlsanity.com/paper/2506.06318/) (2025)
- [Improving the Intent Classification accuracy in Noisy Environment](https://mlsanity.com/paper/2303.06585/) (2023)
- [How much to Dereverberate? Low-Latency Single-Channel Speech Enhancement in Distant Microphone Scenarios](https://mlsanity.com/paper/2505.01338/) (2025)
- [ISAC: An Invertible and Stable Auditory Filter Bank with Customizable Kernels for ML Integration](https://mlsanity.com/paper/2505.07709/) (2025)
- [Advances in Intelligent Hearing Aids: Deep Learning Approaches to Selective Noise Cancellation](https://mlsanity.com/paper/2507.07043/) (2025)
- [A Small-footprint Acoustic Echo Cancellation Solution for Mobile Full-Duplex Speech Interactions](https://mlsanity.com/paper/2508.07561/) (2025)
- [Alternating Approach-Putt Models for Multi-Stage Speech Enhancement](https://mlsanity.com/paper/2508.10436/) (2025)
- [Towards Lightweight Adaptation of Speech Enhancement Models in Real-World Environments](https://mlsanity.com/paper/2603.07471/) (2026)
- [Real-Time Band-Grouped Vocal Denoising Using Sigmoid-Driven Ideal Ratio Masking](https://mlsanity.com/paper/2603.29326/) (2026)
- [QC-GAN: A Parameter-Efficient Quaternion Conformer GAN for High-Fidelity Speech Enhancement](https://mlsanity.com/paper/2606.18611/) (2026)
- [Does per-frame early exit pay? A compute-matched study of dynamic depth for on-device speech enhancement](https://mlsanity.com/paper/2609.29867/) (2026)

## Abstract (arXiv)

Speech enhancement using neural networks is recently receiving large attention in research and being integrated in commercial devices and applications. In this work, we investigate data augmentation techniques for supervised deep learning-based speech enhancement. We show that not only augmenting SNR values to a broader range and a continuous distribution helps to regularize training, but also augmenting the spectral and dynamic level diversity. However, to not degrade training by level augmentation, we propose a modification to signal-based loss functions by applying sequence level normalization. We show in experiments that this normalization overcomes the degradation caused by training on sequences with imbalanced signal levels, when using a level-dependent loss function.