A bit more experimentation with sampling and generation: the following samples are generated by sampling from the embedding space rather than output probabilities. The sound (at least to me) a bit better. Here are a few tracks:
So I’ve been playing with different models for algorithmic generative music for a while as a side project. Lately, the models I’ve been building are starting to sound better than garbage.
I will write at length later on the details of the models, but it’s suffice to say for now that they are close to Deep Dynamic Graphical Models (what else they would be ;-)) trained on a bunch of MIDI files. The algorithm is capable of modelling multiple polyphonic instruments as a whole (meaning multiple instruments playing multiple notes with different values/durations per instrument at the same time) and hence it models more or less the band at its entirety.
Following samples are generated from a heavy metal model with a drummer, a bass guitarist, and two electric guitarists – lead and rhythm. Me and Aida call the band LnH 🙂
I am learning Hadoop and needed to build the plugin for Eclipse. Following this link I managed to build one for Hadoop 1.2.1 and Eclipse Indigo on Kubuntu 12.04 LTS. Feel free to download it and let me know if it worked.
This is exciting. At our computer vision group at London Studio, we have been working on a augmented reality gaming and story-telling PS3 peripheral which is called Wonderbook. Wonderbook is an augmented reality book where the story comes alive on the pages of the book. The first game is a collaboration with J.K. Rowling (author of Harry Potter books) and is titled “Book of Spells”.
It was announced at E3 2012, and here are a couple of early videos and trailers.