Multimodal System for Recommending Music

Silvester Jakša (2015) Multimodal System for Recommending Music. EngD thesis.

Preview

Abstract

Our work discusses development of multimodal music recommendation system. Users of the system are walking on the bridge while using mobile application and the system recommends music according to the walking style, adjusts tempo of the playing song, illuminates the bridge and projects user's positions on the floor. For this particular project we developed electronic boards and used them for all user tracking devices. All electronic boards are merged into common hardware which communicates with Rasberry Pi computer. Walking tempo is detected by mobile application which sends all acquired data to multimedia server together with the hardware data. Multimedia server receives data from local network and projects user's positions and generates music. Music and bridge illumination is generated automatically and is recommended to users according to Moodo dataset. Additional system function is playing piano tones which plays regarding users presence on the bridge and their position. The system evalvation was done using short survey after every use. The results showed that the system presents diversification in users daily routine.

Item Type:

Thesis (EngD thesis)

Keywords:

music recommendation, multimodal perception, proximity sensors, Raspberry Pi, Moodo dataset, mobile application, Moodwalker

Number of Pages:

Language of Content:

Slovenian

Mentor / Comentors:

Name and Surname	ID	Function
prof. dr. Saša Divjak	233	Mentor

Link to COBISS:

http://www.cobiss.si/scripts/cobiss?command=search&base=51012&select=(ID=1536474819)

Institution:

University of Ljubljana

Department:

Faculty of Computer and Information Science

Item ID:

3049

Date Deposited:

07 Sep 2015 14:37

Last Modified:

17 Sep 2015 09:54

URI:

http://eprints.fri.uni-lj.si/id/eprint/3049

Actions (login required)

View Item