Meeting minutes: 2/16/2016

Time: 15:20-15:40

Attendees: Ken, Matthew, Nicole, Chris

New business

1. Training Data for neural nets

2. Goals

3. Architecture - input source processors generate data from MP3 or other audio format files and convert it to data suitable for use at the input layer of a Neural Net Classifier. The NN Classifier (or set of classifiers)  identifies attributes about the musical note being played (what instrument, and so forth) and produces an object with those attributes. The final part uses the objects to produce our usable output - some sort of musical notation TAB(lature) [e.g.: guitar TABs]  or MIDI output. (I think there are tools to produce shet music from MIDI data -- investigate)

5. selected Python as language for now

4. review proposal and agre to submit.

Completed since 2/11/16

All - discuss project and agree to initial goals.

Ken - set up Google drive, granted access to group embers, drafted and uploaded proposal document, started researching MP3 format, and python libraries for processing MP3s.

Chris - coordinated group establishment

Nicole - investigated potential difficulties in project, identified items needing further clarification.

Deliverables

All - review input data format - MP3/WAV - processing in python

Ken - set up blog, rip some training data from personal library, 

Nicole - submit proposal to BBLearn.