Hydranet: A Real-Time Waveform Separation Network

Kaspersen, Esbern Torgard; Kounalakis, Tsampikos; Erkut, Cumhur

doi:10.1109/icassp40776.2020.9053357

Cited by 4 publications

(2 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Thus, an architecture similar to the Wave-U-Net [4] was used, replacing the convolutional bottleneck of the network with a recurrent module. The recurrent path consists of two bidirectional LSTM layers, as also proposed in [10], [18], of 168 units each, and a LeakyReLU activation after the second layer. Also, in order to keep the computational costs low, the number of filters in both the encoder and the decoder were halved compared to the original implementation [4].…”

Section: B Denoising Networkmentioning

confidence: 99%

HTMD-Net: A Hybrid Masking-Denoising Approach to Time-Domain Monaural Singing Voice Separation

Garoufis¹,

Zlatintsi²,

Maragos³

2021

Preprint

View full text Add to dashboard Cite

The advent of deep learning has led to the prevalence of deep neural network architectures for monaural music source separation, with end-to-end approaches that operate directly on the waveform level increasingly receiving research attention. Among these approaches, transformation of the input mixture to a learned latent space, and multiplicative application of a soft mask to the latent mixture, achieves the best performance, but is prone to the introduction of artifacts to the source estimate. To alleviate this problem, in this paper we propose a hybrid time-domain approach, termed the HTMD-Net, combining a lightweight masking component and a denoising module, based on skip connections, in order to refine the source estimated by the masking procedure. Evaluation of our approach in the task of monaural singing voice separation in the musdb18 dataset indicates that our proposed method achieves competitive performance compared to methods based purely on masking when trained under the same conditions, especially regarding the behavior during silent segments, while achieving higher computational efficiency.

show abstract

Section: B Denoising Networkmentioning

confidence: 99%

HTMD-Net: A Hybrid Masking-Denoising Approach to Time-Domain Monaural Singing Voice Separation

Garoufis¹,

Zlatintsi²,

Maragos³

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Specifically, a temporal convolutional network for real time speech enhancement has been proposed in [27] while the latest state-of-the-art performance has been obtained by a real-time variation of Demucs for online speech denoising [6]. In a similar sense real-time music source separation models have been proposed in [9] and a system capable of real-time speech separation from background music has been implemented in [14].…”

Section: Introductionmentioning

confidence: 99%

Compute and memory efficient universal sound source separation

Tzinis,

Wang,

Jiang

et al. 2021

Preprint

View full text Add to dashboard Cite

Recent progress in audio source separation lead by deep learning has enabled many neural network models to provide robust solutions to this fundamental estimation problem. In this study, we provide a family of efficient neural network architectures for general purpose audio source separation while focusing on multiple computational aspects that hinder the application of neural networks in real-world scenarios. The backbone structure of this convolutional network is the SUccessive DOwnsampling and Resampling of Multi-Resolution Features (SuDoRM-RF) as well as their aggregation which is performed through simple one-dimensional convolutions. This mechanism enables our models to obtain high fidelity signal separation in a wide variety of settings where variable number of sources are present and with limited computational resources (e.g. floating point operations, memory footprint, number of parameters and latency). Our experiments show that SuDoRM-RF models perform comparably and even surpass several state-of-the-art benchmarks with significantly higher computational resource requirements. The causal variation of SuDoRM-RF is able to obtain competitive performance in real-time speech separation of around 10dB scale-invariant signal-to-distortion ratio improvement (SI-SDRi) while remaining up to 20 times faster than real-time on a laptop device.

show abstract