D3Net: Densely connected multidilated DenseNet for music source separation

Abstract

Music source separation involves a large input field to model a long-term dependence of an audio signal. Previous convolutional neural network (CNN)-based approaches address the large input field modeling using sequentially down- and up-sampling feature maps or dilated convolution. In this paper, we claim the importance of a rapid growth of a receptive field and a simultaneous modeling of multi-resolution data in a single convolution layer, and propose a novel CNN architecture called densely connected dilated DenseNet (D3Net). D3Net involves a novel multi-dilated convolution that has different dilation factors in a single layer to model different resolutions simultaneously. By combining the multi-dilated convolution with DenseNet architecture, D3Net avoids the aliasing problem that exists when we naively incorporate the dilated convolution in DenseNet. Experimental results on MUSDB18 dataset show that D3Net achieves state-of-the-art performance with an average signal to distortion ratio (SDR) of 6.01 dB.

References in corpus (2)

• Voice Separation with an Unknown Number of Multiple Speakers
• Listening to the World Improves Speech Command Recognition

Cited by in corpus (8)

• Hybrid Spectrogram and Waveform Source Separation
• A Comprehensive Survey of 3D Dense Captioning: Localizing and Describing Objects in 3D Scenes
• Separate but Together: Unsupervised Federated Learning for Speech Enhancement from Non-IID Data
• ACCDOA: Activity-Coupled Cartesian Direction of Arrival Representation for Sound Event Localization and Detection
• Sound Event Localization and Detection Using Activity-Coupled Cartesian DOA Vector and RD3net
• End-to-end lyrics Recognition with Voice to Singing Style Transfer
• Unsupervised Source Separation via Bayesian Inference in the Latent Domain
• Multi-Task Audio Source Separation