Roomba Recon
From Music Hack Day
Contents |
Creator
About the hack
The inspiration for this project is a robot solving a maze. Or cleaning a house. The general idea is to construct coherent playlists that traverse from a start to end song on the Soundcloud with minimal prior knowledge. It is built in Python with CherryPy running the web server. Playlists are constructed using the social connections of soundcloud and measures of track-to-track similarity. A number of similarities measures are used, including timbrel similarity with timbre analysis from the echonest and a tf/idf vector space model formed from a tracks comments and tags. Roomba Recon uses a number of APIs and publicly accessible tools including Soundcloud's python API wrapper, the echonest for audio analysis and gensim for natural language processing. Written and conceived by Ben Fields for the 2010 Music Hackday. Banner was created by Becky Stewart
This project has no affiliation with iRobot or the Roomba autonomous robotic vacuum cleaner, other than an appreciation of its awesome.
What APIs, tools or kit did you use?
- Soundcloud (python api wrapper and js custom player)
- echonest
- python
- cherrypy
- gensim
Images/Screenshots
Anything left to do
better (faster) song to song distance would be nice. Some ajax as well. Perhaps a nicer way to start and end on arbitrary soundcloud tracks then manual URL construction still sorting out deployment. hopefully, will be live instead of locale for demonstration
