Marsyas (Music Analysis, Retrieval and Synthesis for Audio Signals) is an open source software framework for audio processing with specific emphasis on Music Information Retrieval applications. It has been designed and written by George Tzanetakis (gtzan@cs.uvic.ca) with help from students and researchers from around the world. Marsyas has been used for a variety of projects in both academia and industry.
Finalist in the Sourceforge
Community Choice Awards 2009
Featured Project
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.
I am an assistant professor in Computer Science (also cross-listed in Music and Electrical and Computer Engineering) at the University of Victoria in Canada. I received my PhD in Computer Science from Princeton University under the supervision of Dr. Perry Cook. I also worked at Carnegie Mellon University as a PostDoctoral Fellow with Dr. Roger Dannenberg on query-by-humming systems and audio-score alignment and with the Informedia group on multimodal video retrieval and microphone arrays.
I have also consulted with several companies using Marsyas. They include&58 Moodlogic Inc. (audio fingerpriting), All Music Inc, The Netherlands (music-speech classification), and Teligence Communications Inc. (gender classification of voice recordings).
My research deals with all stages of audio content analysis such as feature extraction, segmentation, classification, retrieval and source separation with specific focus on Music Information Retrieval (MIR). I am also an active musician and have studied saxophone performance, music theory and composition.
Cantillion is a web based Flash application that allows ethnomusiologists around the world to explore tropes and gestures from partially annotated chant traditions from around the world. We are working on using Marsyas to generate the pitch contours and to perform Dynamic Time Warping similiarity calculations between different tropes.
Featured Video
KeyBoard Controlled Assistive Music Browser
The goal of this project is to develop Assistive Music Browser Software that will enable users with severe disabilities to navigate through their own collections of music via a variety of input methods. We extract relevant features and analyze the audio data within a collection of music (of essentially unlimited size) and map the songs onto a 2-D space, called a self-organizing map, based on similarity. Continual audio feedback is provided to the user as they search their collection to aide in navigation and selection. http://www.canassist.ca