Multi-Scale Chroma n-Gram Indexing for Cover Song Identification

Jin S. SEO  

Publication
IEICE TRANSACTIONS on Information and Systems   Vol.E103-D   No.1   pp.59-62
Publication Date: 2020/01/01
Online ISSN: 1745-1361
DOI: 10.1587/transinf.2019MUL0001
Type of Manuscript: Special Section LETTER (Special Section on Enriched Multimedia — Application of Multimedia Technology and Its Security —)
Category: 
Keyword: 
multi-scale indexing,  cover song,  music retrieval,  n-gram,  chromagram,  

Full Text: PDF(229.9KB)>>
Buy this Article




Summary: 
To enhance cover song identification accuracy on a large-size music archive, a song-level feature summarization method is proposed by using multi-scale representation. The chroma n-grams are extracted in multiple scales to cope with both global and local tempo changes. We derive index from the extracted n-grams by clustering to reduce storage and computation for DB search. Experiments on the widely used music datasets confirmed that the proposed method achieves the state-of-the-art accuracy while reducing cost for cover song search.