Deep neural network based instrument extraction from music

By Stefan Uhlich, Franck Giron, Yuki Mitsufuji. 2015

131 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition, Signal Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

In Signal Processing

Citations in the corpus

Music Source Separation in the Waveform Domain (2019, background): “, 2014], followed by works on music using simple fully connected networks over few spectrogram frames [Uhlich et al., 2015], LSTMs [Uhlich et al.”
Multichannel music separation with deep neural networks (2016, unknown): “The DNNs can be used either to predict the source spectrograms [5]–[8] whose ratio yields a time-frequency (TF) mask or directly to predict a TF mask [9]–[13].”
Multi-Scale multi-band densenets for audio source separation (2017, methodology): “• FNN [11]: The source spectra was estimated by feed forward fully connected DNN trained with an additional dataset (Med-leyDB [28]).”
The 2018 Signal Separation Evaluation Campaign (2018, methodology): “ditional data These methods are data-driven and exploit only the training data for MUSDB18 to learn the models. They are: RGT1-2 [23], STL, HEL1 [10], MDL1 [17], MDLT [16], JIY1-3 [11], WK [36], UHL1 [27], UHL2 [28]. With additional data These methods are also data-driven, and exploit additional tra”
D3Net: Densely connected multidilated DenseNet for music source separation (2020, background): “Many types of neural network architecture have been proposed, including feedforward fully connected networks (FNNs) [1, 2], recurrent neural networks (RNNs) [3], convolutional neural networks (CNNs) [4–9] and their combinations [10,11].”
Mmdenselstm: An Efficient Combination of Convolutional and Recurrent Neural Networks for Audio Source Separation (2018, methodology): “In [10, 11], a standard feedforward fully connected network (FNN) was used to estimate source spectra.”
Multichannel Audio Source Separation With Deep Neural Networks (2016, background): “The estimated source signal is then obtained as the product of the input mixture signal and the estimated time-frequency mask.”
Open-Unmix - A Reference Implementation for Music Source Separation (2019, background): “No open-source system is available today that matches the performance of the best system proposed more than four years ago by (Uhlich, Giron, & Mitsufuji, 2015).”
Monoaural Audio Source Separation Using Deep Convolutional Neural Networks (2017, methodology): “[16] propose another method which takes multiple frames of the magnitude spectrogram of a mixture as input and consists of only fully connected layers.”