2019: Teaching Robots how to Drum

or Auld Bot Syne

30 Dec 2019 - atrodo - Song: Auld Lang Syne by Relient K

As 2019 closes, I wanted to take a moment and reflect on everything I’ve done this year with Teaching Robots. The first commit I made was in late 2018, so most of the development has been done this year. Looking back, I can see that I’ve done a lot of coding and discovery this year.

In the past year, I accomplished a lot of things:

Made around 100 commits
Made about 15 blog posts
Created a Neural Network from scratch
Learned x86 SIMD/intrinsics in C
Used SIMD to speed up my Neural Network code
Generated very rudimentary drum beats with Neural Networks
Implemented a genetic algorithm to create new Neural Network variants
Figured out how to write fitness functions to test music
Figured out how not to write fitness functions
Connected the Neural Network up to generate a sound file
Learned a lot about what Neural Networks can and cannot do
Created an IRC bot to aid in managing the various data types
Made a streaming process to reduce the feedback loop
Discovered how DFT and FFT work, in theory and in practice

Not the least of all of this was the four month stint working on FFT and Melody Extraction. A noticeable portion of those four months was just in the two posts.

Looking through that list, even with the time spent working just one one project, I feel good about what I’ve accomplished. I’ve done a lot this year, and I am hoping that soon I’ll be able to start creating something musical from all of this work and start refining everything I’ve created so far.