Tambr is a software for creating music from text.
Tambr runs a topic extraction algorithm to get topics from a text, then it finds synthesizers that sound the most like the topics by using a search engine implemented from human-labeled training data and on word-embeddings extracted from Google Word2Vec
There is no unified GUI. In order for Tambr to be usable for all, a UI is necessary and is the most important part to getting it off the ground as a standalone program.
Tambr was created by and is maintained by Jessie Salas, Machine learning software engineer and musician.
##How To Contribute The codebase will be progressively organized and maintained by the creator. If you are interested in collaborating, feel free to contact me for more information about how things work. The files are looseley stringed together at this point, and a unified interface is of the highest priority to move forward.