Published August 8, 2017 | Version v1
Software Open

Trained Models for "A Recurrent Encoder-Decoder Approach With Skip-Filtering Connections For Monaural Singing Voice Separation"

Creators

  • 1. Fraunhofer IDMT, Ilmenau, Germany.
  • 2. Audio Research Group, Dept. of Signal Processing, Tampere University of Technology, Tampere, Finland.
  • 3. Technical University of Ilmenau, Ilmenau, Germany.

Description

Support material (binary files) for the following work: S.I. Mimilakis, K. Drossos, T. Virtanen, G. Schuller, "A Recurrent Encoder-Decoder Approach With Skip-Filtering Connections For Monaural Singing Voice Separation", accepted for presentation at the 2017 IEEE International Workshop on Machine Learning for Signal Processing, September 25–28, 2017, Tokyo, Japan.

To be used here: https://github.com/Js-Mim/mlsp2017_svsep_skipfilt/

Files

BiGRU_models.zip

Files (411.0 MB)

Name Size Download all
md5:80dc120fc769a079e9b68bf7c59f66c6
411.0 MB Preview Download

Additional details

Funding

MacSeNet – Machine Sensing Training Network 642685
European Commission
EVERYSOUND – Computational Analysis of Everyday Soundscapes 637422
European Commission