Yahoo Research

ORBIT Project (BBC Research)

VISNET I & II EU FP6 NoE Projects

INESC Porto

Teligence

Moodlogic

last.fm

MusicEmo - Eric Yang, NTU, Taiwan

Musicream

Musie Mood

---

Yahoo Research

Yahoo Research and the Yahoo Media Group have been using Marsyas to analyze songs in our 2-million song database. We are interested in both how the songs are related to each other, so we can find similar songs, and how they differ, so we can characterize people's musical interests. Marsyas provides us with an easy-to-use platform that computes many of the features we think are useful. But more importantly, Marsyas is a common platform for acoustic signal processing, so we can report results that are easy for other research labs to replicate. - Malcolm Slaney
http://research.yahoo.com/~malcolm

ORBIT Project (BBC Research)

ORBIT - “Object Re-configurable Broadcast Infrastructure Trial”, a contract pilot project between the BBC R&D (http://www.bbc.co.uk/rd/index.shtml) and INESC Porto (http://www.inescporto.pt): development of automatic audio segmentation and classification tools using Marsyas (September 2001 ~ September 2002).
http://www.bbc.co.uk/orbit/index.shtml

VISNET I & II EU FP6 NoE Projects

VISNET I & II - “NETworked audio VISual media technologies” (FP6-2002-IST-1 and FP6-2005-IST-41 European Union Network of Excellence Projects, respectively): Marsyas is used at Audio and Music analysis work packages by some of the partners (December 2003 ~ June 2009).
http://www.visnet-noe.org/

INESC Porto

INESC Porto - Institute for Systems and Computer Engineering of Porto - is a private non-profit association, recognized as Public Interest Institution, that has been recently appointed as Associated Laboratory. It is located in Porto, Portugal. At the Multimedia and Telecommunications Unit (UTM), INESC Porto researchers have used and contributed to Marsyas development in audio, video and multimodal analysis and processing. More info can be found at: http://www.inescporto.pt/~lmartins/ http://telecom.inescporto.pt/~lfpt/main/pmwiki.php?n=Main.MarsyasX
http://www.inescporto.pt

Teligence

Marsyas was used to design and develop an in-house tool for gender classification (male/female/silence) for voice messages. The system achieves classification of approximately 90% and is running on 10 hubs processing about 25000 recordings (1-2 minutes each) per day. The project was initiated by Paul Snider with consulting by George Tzanetakis.
http://www.teligence.net/

Moodlogic

Marsyas was used to design and prototype the audio fingerprinting technology used to link user files to metadata and fix ID3 tags by the Moodlogic client. The fingerprint is small (about 300 bytes/file), is fast to compute, and matching is performed to a database of 1.5 million songs)
http://www.moodlogic.com/

last.fm

"At last.fm we used Marsyas to design and validate prototypes, and to quickly test ideas. The latest version dramatically improved its quality and performance - if you are doing serious MIR your should definitely give it a try!" - Norman Casagrande, Head of Music Research
http://www.last.fm

MusicEmo - Eric Yang, NTU, Taiwan

We identified and analyze three critical issues of music emotion recognition: the subjectivity of emotion perception, the ambiguity of categorical emotion models, and the semantic gap between low-level audio signals to high-level emotion. Some results of the previous two issues have been published. To see more detail, please see the website. Marsyas is used in our system for feature extraction. Our experiments show the strength of the features in Marsyas.
http://mpac.ee.ntu.edu.tw/~yihsuan/

Musicream

Musicream is a novel music playback interface that lets users unexpectedly come across musical pieces they like. It facilitates active, flexible, and unexpected browsing. For example, the "similarity-based" sticking function enables user to easily pick out and listen to similar pieces from a streaming "flow" of music. Marsyas is used to automatically extract a single feature vector that characterizes the content of a particular music piece. That vector is used for color visualization of the audio content as well as to support the "similarity-based" sticking function. Marsyas provided an easy way to extract audio content information and enabled us to concentrate on designing and developing the user interface. Masataka Goto Senior Research Scientist, AIST, Japan
http://staff.aist.go.jp/m.goto/Musicream/

Musie Mood

our goal is to develop an integrated system to visualize and query large music libraries. The layout is controlled by the user and similarity of songs is measured in perceptual terms. We use Marsyas to extract structural features which have perceptual interpretation (e.g. tempo, loudness, beat strength, etc...). - Vladimir G. Kim, Steven Bergner, Torsten Möller. GrUVi lab, Simon Fraser Univeristy, Canada.
http://gruvi.cs.sfu.ca/researchProject.php?s=373