Abstract | ||
---|---|---|
We introduce SongVis, a visualization that represents music's semantic descriptors. SongVis uses emojis, colors, lines and shapes to embody the semantic content of a song. It aims to aid users on tasks related to exploration/browsing of music libraries and at queries for similar tracks based on visual characteristics. We first collected the descriptors after surveying papers on the topic of "music visualization", and used a questionnaire to rank the terms by consulting the public. Then, the features: mood, danceability, tempo, music genre and instrument, were extracted using state-of-the-art music information retrieval algorithms and we considered their visuals. Then, we discuss potential improvements to be made. With SongVis we perform a step forward towards visually representing semantic descriptors and expect that the music information research, the information visualization fields and the public can benefit from this work. |
Year | DOI | Venue |
---|---|---|
2019 | 10.1109/IV.2019.00066 | 2019 23rd International Conference Information Visualisation (IV) |
Keywords | DocType | ISSN |
music visualization,information visualization,semantic descriptors | Conference | 1550-6037 |
ISBN | Citations | PageRank |
978-1-7281-2839-9 | 0 | 0.34 |
References | Authors | |
9 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Hugo Lima | 1 | 0 | 0.34 |
Carlos Santos | 2 | 0 | 0.34 |
Bianchi Serique Meiguins | 3 | 56 | 28.03 |