speech-2

Multi-Speaker Source Separation

3260 papers • 126 benchmarks • 313 datasets

This task has no description! Would you like to contribute one?

(Image credit: Papersgraph)

Benchmarks

These leaderboards are used to track progress in multi-speaker-source-separation-2

Trend

Dataset

Best Model

Actions

No benchmarks available.

Libraries

i

Use these libraries to find multi-speaker-source-separation-2 models and implementations

JeroenZegers/Nabu-MSSS

3 papers 18

Datasets

No datasets available.

Subtasks

No subtasks available.

Most implemented papers

Directional Sparse Filtering Using Weighted Lehmer Mean for Blind Separation of Unbalanced Speech Mixtures

Karn N. Watcharasupat, Anh H. T. Nguyen, Andy W. H. Khong, Ching-Hui Ooi•Fri Jan 29 2021

An algorithm based on the directional sparse filtering (DSF) framework that utilizes the Lehmer mean with learnable weights to adaptively account for source imbalance is proposed.

1

Content

0

Paper Graph

Memory Time Span in LSTMs for Multi-Speaker Source Separation

H. V. Hamme, Jeroen Zegers•Thu Aug 23 2018

This paper analyzed how recurrent neural network (RNNs) cope with temporal dependencies by determining the relevant memory time span in a long short-term memory (LSTM) cell by leaking the state variable with a controlled lifetime and evaluating the task performance.

5 0

Paper Graph

Multi-Scenario Deep Learning for Multi-Speaker Source Separation

H. V. Hamme, Jeroen Zegers•Thu Apr 12 2018

This work has shown that data of a specific scenario is relevant for solving another scenario, and concluded that a single model, trained on different scenarios is capable of matching performance of scenario specific models.

3 0

Paper Graph

Unsupervised Deep Clustering for Source Separation: Direct Learning from Mixtures Using Spatial Information

Paris Smaragdis, Efthymios Tzinis, Shrikant Venkataramani•Sun Nov 04 2018

A deep clustering approach is used which trains on multichannel mixtures and learns to project spectrogram bins to source clusters that correlate with various spatial features, and shows that this system is capable of performing sound separation on monophonic inputs, despite having learned how to do so using multi-channel recordings.

53 0

Paper Graph

CNN-LSTM models for Multi-Speaker Source Separation using Bayesian Hyper Parameter Optimization

H. V. Hamme, Jeroen Zegers•Sat Sep 14 2019

A Bayesian hyper parameter optimization technique is used and it is found that the parallel CNN-LSTM outperforms the LSTM-only and CNN-only model.

8 0

Paper Graph

Adding a benchmark result helps the community track progress.