@lachlanjcIMA @ NYU

CC – Week 7: Sound and Video – Log

This week I'm using the p5.speech library to make something fun with web speech recognition.

(Note: requires Chrome or another browser that supports the Web Speech API.)

Open Demo

How this works

let speech = new p5.SpeechRec('en-US', parseResult)
speech.continuous = true
speech.interimResults = false
function setup() {
createCanvas(window.innerWidth, window.innerHeight)
background(255)
fill(25)
textSize(48)
textAlign(CENTER)
textStyle(BOLDITALIC)
textFont('"Avenir Next", system-ui, sans-serif')
text('SAY A COLOR', width / 2, height / 2)
speech.start()
}
function draw() {
// Where we’re going we don’t need drawing
}
const colors = []
function parseResult() {
if (speech.resultValue) {
const color = speech.resultString.split(' ').pop().toUpperCase()
colors.push(color)
background(color)
text(color, width / 2, height / 2)
console.log(colors)
}
}
  1. Loading the p5.speech library, in the HTML
  2. Initializing web speech for English, using continuous (as opposed to one-time) dictation
  3. Initializing a full-screen canvas
  4. Drawing some basic instructions on the screen
  5. Whenever the speech recognition returns a result, it runs a function that
    1. Gets the last word of the string, if it's a multi-word phrase
    2. Saves it to an array, primarily for debugging
    3. Sets the background color
    4. Updates the centered onscreen text with the new color name

Voilà! Check out the code on P5.

Lachlan's avatar