eduzhai > Physical Sciences > Physics Sciences >

Low-Rank Sparse Representation with Pre-Learned Dictionaries and Side Information for Singing Voice Separation

  • Save

... pages left unread,continue reading

Document pages: 9 pages

Abstract: At present, although the human speech separation hasachieved fruitful results, it is not ideal for the separation of singing andaccompaniment. Based on low-rank and sparse optimization theory, in this paper,we propose a new singing voice separation algorithm called Low-rank, SparseRepresentation with pre-learned dictionaries and side Information (LSRi). Thealgorithm incorporates both the vocal and instrumental spectrograms as sparsematrix and low-rank matrix, meanwhile combines pre-learning dictionary and thereconstructed voice spectrogram form the annotation. Evaluations on the iKaladataset show that the proposed methods are effective and efficient for singingvoice separation.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...