# Dense CNN with Self-Attention for Time-Domain Speech Enhancement

arXiv: https://arxiv.org/abs/2009.01941 · HTML: https://mlsanity.com/paper/2009.01941/

## Key facts

- Cited by 16 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 3 Sep 2020.

## Built on by (most cited first, 16 of 16)

- [Personalized Speech Enhancement: New Models and Comprehensive Evaluation](https://mlsanity.com/paper/2110.09625/) (2021)
- [RemixIT: Continual self-training of speech enhancement models via bootstrapped remixing](https://mlsanity.com/paper/2202.08862/) (2022)
- [Self-Supervised Learning for Speech Enhancement through Synthesis](https://mlsanity.com/paper/2211.02542/) (2022)
- [Wav2code: Restore Clean Speech Representations via Codebook Lookup for Noise-Robust ASR](https://mlsanity.com/paper/2304.04974/) (2023)
- [Continual self-training with bootstrapped remixing for speech enhancement](https://mlsanity.com/paper/2110.10103/) (2021)
- [SkipConvGAN: Monaural Speech Dereverberation using Generative Adversarial Networks via Complex Time-Frequency Masking](https://mlsanity.com/paper/2211.12623/) (2022)
- [BASEN: Time-Domain Brain-Assisted Speech Enhancement Network with Convolutional Cross Attention in Multi-talker Conditions](https://mlsanity.com/paper/2305.09994/) (2023)
- [Modulating State Space Model with SlowFast Framework for Compute-Efficient Ultra Low-Latency Speech Enhancement](https://mlsanity.com/paper/2411.02019/) (2024)
- [Multichannel-to-Multichannel Target Sound Extraction Using Direction and Timestamp Clues](https://mlsanity.com/paper/2409.12415/) (2024)
- [MambAttention: Mamba with Multi-Head Attention for Generalizable Single-Channel Speech Enhancement](https://mlsanity.com/paper/2507.00966/) (2025)
- [A Training and Inference Strategy Using Noisy and Enhanced Speech as Target for Speech Enhancement without Clean Speech](https://mlsanity.com/paper/2210.15368/) (2022)
- [Heterogeneous Space Fusion and Dual-Dimension Attention: A New Paradigm for Speech Enhancement](https://mlsanity.com/paper/2408.06911/) (2024)
- [SEOFP-NET: Compression and Acceleration of Deep Neural Networks for Speech Enhancement Using Sign-Exponent-Only Floating-Points](https://mlsanity.com/paper/2111.04436/) (2021)
- [Objective and subjective evaluation of speech enhancement methods in the UDASE task of the 7th CHiME challenge](https://mlsanity.com/paper/2402.01413/) (2024)
- [Masked Autoencoders as Universal Speech Enhancer](https://mlsanity.com/paper/2602.02413/) (2026)
- [The Costs of Reproducibility in Music Separation Research: a Replication of Band-Split RNN](https://mlsanity.com/paper/2603.09187/) (2026)

## Abstract (arXiv)

Speech enhancement in the time domain is becoming increasingly popular in recent years, due to its capability to jointly enhance both the magnitude and the phase of speech. In this work, we propose a dense convolutional network (DCN) with self-attention for speech enhancement in the time domain. DCN is an encoder and decoder based architecture with skip connections. Each layer in the encoder and the decoder comprises a dense block and an attention module. Dense blocks and attention modules help in feature extraction using a combination of feature reuse, increased network depth, and maximum context aggregation. Furthermore, we reveal previously unknown problems with a loss based on the spectral magnitude of enhanced speech. To alleviate these problems, we propose a novel loss based on magnitudes of enhanced speech and a predicted noise. Even though the proposed loss is based on magnitudes only, a constraint imposed by noise prediction ensures that the loss enhances both magnitude and phase. Experimental results demonstrate that DCN trained with the proposed loss substantially outperforms other state-of-the-art approaches to causal and non-causal speech enhancement.