Skip to content
This repository has been archived by the owner on Aug 27, 2020. It is now read-only.

Latest commit

 

History

History

07_rnn

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 
 
 
 
 

Recurrent Neural Network (RNN)

Session A: SketchRNN (RNN with Pre-Trained Models 1)

Objectives:

  • Learn about Sequential Data and Markov Chains
  • Learn about Recurrent Neural Network architectures
  • Become familiar with use cases for RNNs
  • Learn ml5’s RNN functions and underlying pre-trained models

Markov Chains

RNNs

Related Projects:

RDP (Ramer-Douglas-Peucker) Algorithm

SketchRNN Examples

Session B: Magenta.js (RNN with Pre-Trained Models 2)

Objectives

  • Learn distinction between LSTM and RNN.
  • Become familiar with Magenta.js and RNN models for generating music.
  • Understand concept of “temperature” as it relates to RNNs.

Neural Networks and Music

Resources

Related Projects

MusicRNN Examples

LSTM

Assignment 6 Due Sunday October 20 at 6pm

Reading

Instructions

Use ml5.sketchRNN() and/or magenta.js to generate visuals and/or sound in an interactive p5.js sketch. You can invent your own creative idea or here are some suggested exercises:

  • As suggested by @hardmaru, implement the RDP Line Simplification algorithm in the Interactive Drawings with SketchRNN example. Does the program perform "better" or "worse" (what does that even mean)? What is the optimal epsilon value? (A video tutorial about RDP algorithm will be shared via the mailing list.)
  • Consider the interaction design of Interactive Drawings with SketchRNN example? How can you improve upon the user experience? Can you craft a story around the interaction or make it into a game?
  • Visualize the melody generated with MelodyRNN. magenta.js has built in functions for doing this that you can experiment with or try drawing your own visualization with p5.js.
  • Create your own version of an "AI Duet", can you build a user interaction around playing a melody that the ML model continues?

Document your coding exercise and how the readings informed your work in a blog post and link from the homework wiki page.