We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.LG

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Machine Learning

Title: Automatic Music Highlight Extraction using Convolutional Recurrent Attention Networks

Abstract: Music highlights are valuable contents for music services. Most methods focused on low-level signal features. We propose a method for extracting highlights using high-level features from convolutional recurrent attention networks (CRAN). CRAN utilizes convolution and recurrent layers for sequential learning with an attention mechanism. The attention allows CRAN to capture significant snippets for distinguishing between genres, thus being used as a high-level feature. CRAN was evaluated on over 32,000 popular tracks in Korea for two months. Experimental results show our method outperforms three baseline methods through quantitative and qualitative evaluations. Also, we analyze the effects of attention and sequence information on performance.
Subjects: Machine Learning (cs.LG); Multimedia (cs.MM); Sound (cs.SD); Machine Learning (stat.ML)
Cite as: arXiv:1712.05901 [cs.LG]
  (or arXiv:1712.05901v1 [cs.LG] for this version)

Submission history

From: Jung Woo Ha [view email]
[v1] Sat, 16 Dec 2017 04:27:36 GMT (177kb,D)

Link back to: arXiv, form interface, contact.