1: Interactive Real-Time Gesture Data

Continue working on your sketch from Assignment 8. What are your next steps? Maybe try breaking it into multiple sketches for saving / loading the trained model? Or try a regression model instead of classification?

**https://editor.p5js.org/tl3417/sketches/hmg3Ec_Xo**

• i tried to connect handpose training with my oscillator
• but it didn’t work
  • potential problems:
    1. ml5 script didn’t work
    2. not sure what the handpose classification can control
• my idea:
  • use different handpose to trigger either different sound effects
  • or draw different shapes on the sketch

2: Working with Quick Draw, DoodleNet

Build off of one of the code examples (or invent your own) to develop you own creative use of Quick, Draw data.

• Classifying Drawings with ml5's DoodleNet: output in DOM element -- Can you make this one work with webcam input instead of canvas?
• Classifying Drawings with ml5's DoodleNet: output in separate canvas
• Can you try your own doodleNet with a fewer classes? demo, code, get doodleNet dataset, download .npy files.

3: Other Data Sources

Try loading JSON or tabular data into a p5.js sketch and train a model. Some suggestions:

• Augment Lydia Jessup's 311 Calls ml5.js example. You could add an additional input field, customize the interface, or change other parameters of ml5.neuralNetwork().
• Train a machine learning model in ml5.js with the dataset you picked for part 1 of the assignment.

Complete a blog post with your dataset report and code exercise. Link from the homework wiki.