Complex Image Generation SwinTransformer Network for Audio Denoising (2023-08-20T00:00:00.000000Z)

TL;DR

A complex image generation SwinTransformer network is developed and structure similarity and detailed loss functions are imposed to generate high-quality images and an SDR loss is developed to minimize the difference between denoised and clean audios.

Abstract

Achieving high-performance audio denoising is still a challenging task in real-world applications. Existing time-frequency methods often ignore the quality of generated frequency domain images. This paper converts the audio denoising problem into an image generation task. We first develop a complex image generation SwinTransformer network to capture more information from the complex Fourier domain. We then impose structure similarity and detailed loss functions to generate high-quality images and develop an SDR loss to minimize the difference between denoised and clean audios. Extensive experiments on two benchmark datasets demonstrate that our proposed model is better than state-of-the-art methods.

Authors

Youshan Zhang

2 papers

TL;DR

Abstract

Authors

References36 items

BirdSoundsDenoising: Deep Visual Audio Denoising for Bird Sounds

Implementing Transformer Architectures for Audio Source Separation

Perception-guided generative adversarial network for end-to-end speech enhancement

MANNER: Multi-View Attention Network For Noise Erasure

A Two-Stage U-Net for High-Fidelity Denoising of Historical Recordings

Speech Denoising in the Waveform Domain With Self-Attention

Lung sound signal denoising using discrete wavelet transform and artificial neural network

S-DCCRN: Super Wide Band DCCRN with Learnable Complex Feature for Speech Enhancement

Mixed Transformer U-Net for Medical Image Segmentation

Heart sound classification based on log Mel-frequency spectral coefficients features and convolutional neural networks

Segmenter: Transformer for Semantic Segmentation

Speech Denoising without Clean Training Data: a Noise2Noise Approach

MetricGAN+: An Improved Version of MetricGAN for Speech Enhancement

TSTNN: Two-Stage Transformer Based Neural Network for Speech Enhancement in the Time Domain

Speech enhancement with weakly labelled data from AudioSet

DCCRGAN: Deep Complex Convolution Recurrent Generator Adversarial Network for Speech Enhancement

Speech enhancement using progressive learning-based convolutional recurrent neural network

Performance Study of a Convolutional Time-Domain Audio Separation Network for Real-Time Speech Denoising

CP-GAN: Context Pyramid Generative Adversarial Network for Speech Enhancement

PHASEN: A Phase-and-Harmonics-Aware Speech Enhancement Network

Phase-aware Speech Enhancement with Deep Complex U-Net

A Convolutional Recurrent Neural Network for Real-Time Speech Enhancement

Deep Denoising for Hearing Aid Applications

Convolutional-Recurrent Neural Networks for Speech Enhancement

Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation

Noisy speech database for training speech enhancement algorithms and TTS models

Deep Recurrent Convolutional Neural Network: Improving Performance For Speech Recognition

A Fully Convolutional Neural Network for Speech Enhancement

Long Short-Term Memory for Speaker Generalization in Supervised Speech Separation

SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation

U-Net: Convolutional Networks for Biomedical Image Segmentation

Image quality assessment: from error visibility to structural similarity

Denoising Speech Based on Deep Learning and Wavelet Decomposition

Swin Transformer: Hierarchical Vision Transformer using Shifted Windows

University of Edinburgh. School of Informatics. Centre for Speech Technology Research (CSTR

Recurrent Neural Networks for Noise Reduction in Robust ASR

Field of Study

Journal Information

Name

Volume

Venue Information

Name

Type

URL

Alternate Names