ICASSP 2013

Title

"Interactive Refinement of Supervised and Semi-Supervised Sound Source Separation Estimates."
N. J. Bryan, G. J. Mysore
IEEE International Conference on Acoustics, Speech, and Signal Processing, Vancouver, Canada. May 2013.
(paper)

Abstract

We propose an interactive refinement method for supervised and semi-supervised single-channel source separation. The reŽnement method allows end-users to provide feedback to the separation process by painting on spectrogram displays of intermediate output results. The time-frequency annotations are then used to update the separation estimates and iteratively refine the results. The initial separation is performed using probabilistic latent component analysis and is then extended to incorporate the painting annotations using linear grouping expectation constraints via the framework of posterior regularization. Using a prototype user-interface, we show that the method is able to perform high-quality separation with minimal user-interaction.

Sound Examples



Example

No Interaction

Interaction
Cell phone + Speech Semi-supervised (source 1, source 2) Semi-supervised (source 1, source 2)
(source 1, source 2, mix) Supervised (source 1, source 2) Supervised (source 1, source 2)
-----------------------------------
Drum + Bass Semi-supervised (source 1, source 2) Semi-supervised (source 1, source 2)
(source 1, source 2, mix) Supervised (source 1, source 2) Supervised (source 1, source 2)
-----------------------------------
Orchestra + Cough Semi-supervised (source 1, source 2) Semi-supervised (source 1, source 2)
(source 1, source 2, mix) Supervised (source 1, source 2) Supervised (source 1, source 2)
-----------------------------------
Piano + Wrong Note Semi-supervised (source 1, source 2) Semi-supervised (source 1, source 2)
(source 1, source 2, mix) Supervised (source 1, source 2) Supervised (source 1, source 2)
-----------------------------------
Siren + Speech Semi-supervised (source 1, source 2) Semi-supervised (source 1, source 2)
(source 1, source 2, mix) Supervised (source 1, source 2) Supervised (source 1, source 2)
-----------------------------------

Video Demonstrations









homepage