change nn -> classifier, comments about sample data

ml5js · Nov 4, 2020 · 3418909 · 3418909
1 parent 207c5c2
commit 3418909
Showing 1 changed file with 10 additions and 8 deletions.
diff --git a/examples/p5js/NeuralNetwork/NeuralNetwork_ImageClassifier_Video/sketch.js b/examples/p5js/NeuralNetwork/NeuralNetwork_ImageClassifier_Video/sketch.js
@@ -8,7 +8,7 @@ ml5 Example
 Image classification using Convolutional Neural Network
 This example uses a callback pattern to create the classifier
 === */
-let nn;
+let classifier;
 const IMAGE_WIDTH = 64;
 const IMAGE_HEIGHT = 64;
 const IMAGE_CHANNELS = 4;
@@ -40,8 +40,10 @@ function setup() {
  };
 
  // construct the neural network
- nn = ml5.neuralNetwork(options);
- // nn.loadData('daytime_nightime.json', train);
+ classifier = ml5.neuralNetwork(options);
+
+ // An option to load sample data for quick testing
+ // classifier.loadData('daytime_nightime.json', train);
 }
 
 function draw() {
@@ -50,7 +52,7 @@ function draw() {
 
 function addData() {
  console.log('adding data', labelInput.value());
- nn.addData({ image: video }, { label: labelInput.value() });
+ classifier.addData({ image: video }, { label: labelInput.value() });
 }
 
 function train() {
@@ -59,13 +61,13 @@ function train() {
  epochs: 4,
  };
 
- nn.normalizeData();
- nn.train(TRAINING_OPTIONS, finishedTraining);
+ classifier.normalizeData();
+ classifier.train(TRAINING_OPTIONS, finishedTraining);
 }
 
 function finishedTraining() {
  console.log('finished training');
- nn.classify([video], gotResults);
+ classifier.classify([video], gotResults);
 }
 
 function gotResults(err, result) {
@@ -74,5 +76,5 @@ function gotResults(err, result) {
  return;
  }
  resultLabel.html(`${result[0].label}`);
- nn.classify([video], gotResults);
+ classifier.classify([video], gotResults);
 }