//***********************************************nn.js ****************************************************************************************************
//**********************************rewrite the ActivationFunction() functiong and some activation function **************************************
var showMatrix = !1;
class ActivationFunction {
constructor(t, e, i = !1) {
this.func = t, this.dfunc = e, this.use_X_values = i
}
}
let sigmoid = new ActivationFunction(t => 1 / (1 + Math.exp(-t)), t => t * (1 - t)),
tanh = new ActivationFunction(t => Math.tanh(t), t => 1 - t * t),
arctan = new ActivationFunction(t => Math.atan(t), t => 1 / (t * t + 1), use_X_values = !0),
softsign = new ActivationFunction(t => t / (1 + Math.abs(t)), t => 1 / Math.pow(Math.abs(t) + 1, 2), use_X_values = !0),
relu = new ActivationFunction(t => t < 0 ? 0 : t, t => t < 0 ? 0 : 1, use_X_values = !0),
leaky_relu = new ActivationFunction(t => t < 0 ? .01 * t : t, t => t < 0 ? .01 : 1, use_X_values = !0),
softplus = new ActivationFunction(t => Math.log(1 + Math.exp(t)), t => 1 / (1 + Math.exp(-t)), use_X_values = !0),
gaussian = new ActivationFunction(t => Math.exp(t * t * -1), t => -2 * t * Math.exp(t * t * -1), use_X_values = !0);
//*************************************************************************************************Yuwei's code end *********************************8
class NeuralNetwork {
/*
* if first argument is a NeuralNetwork the constructor clones it
* USAGE: cloned_nn = new NeuralNetwork(to_clone_nn);
*/
constructor(in_nodes, hid_nodes, out_nodes) {
if (in_nodes instanceof NeuralNetwork) {
let a = in_nodes;
this.input_nodes = a.input_nodes;
this.hidden_nodes = a.hidden_nodes;
this.output_nodes = a.output_nodes;
this.weights_ih = a.weights_ih.copy();
this.weights_ho = a.weights_ho.copy();
this.bias_h = a.bias_h.copy();
this.bias_o = a.bias_o.copy();
} else {
this.input_nodes = in_nodes;
this.hidden_nodes = hid_nodes;
this.output_nodes = out_nodes;
this.weights_ih = new Matrix(this.hidden_nodes, this.input_nodes);
this.weights_ho = new Matrix(this.output_nodes, this.hidden_nodes);
this.weights_ih.randomize();
this.weights_ho.randomize();
this.bias_h = new Matrix(this.hidden_nodes, 1);
this.bias_o = new Matrix(this.output_nodes, 1);
this.bias_h.randomize();
this.bias_o.randomize();
}
// TODO: copy these as well
this.setLearningRate();
this.setActivationFunction();
}
predict(input_array) {
// Generating the Hidden Outputs
let inputs = Matrix.fromArray(input_array);
let hidden = Matrix.multiply(this.weights_ih, inputs);
hidden.add(this.bias_h);
// activation function!
hidden.map(this.activation_function.func);
// Generating the output's output!
let output = Matrix.multiply(this.weights_ho, hidden);
output.add(this.bias_o);
output.map(this.activation_function.func);
// Sending back to the caller!
return output.toArray();
}
setLearningRate(learning_rate = 0.1) {
this.learning_rate = learning_rate;
}
//////////////////////////////
// setActivationFunction(func = sigmoid)
setActivationFunction(func = sigmoid){
this.activation_function = func;
}
train(input_array, target_array) {
// Generating the Hidden Outputs
let inputs = Matrix.fromArray(input_array);
let hidden = Matrix.multiply(this.weights_ih, inputs);
hidden.add(this.bias_h);
// activation function!
hidden.map(this.activation_function.func);
// Generating the output's output!
let outputs = Matrix.multiply(this.weights_ho, hidden);
outputs.add(this.bias_o);
outputs.map(this.activation_function.func);
// Convert array to matrix object
let targets = Matrix.fromArray(target_array);
// Calculate the error
// ERROR = TARGETS - OUTPUTS
let output_errors = Matrix.subtract(targets, outputs);
// let gradient = outputs * (1 - outputs);
// Calculate gradient
let gradients = Matrix.map(outputs, this.activation_function.dfunc);
gradients.multiply(output_errors);
gradients.multiply(this.learning_rate);
// Calculate deltas
let hidden_T = Matrix.transpose(hidden);
let weight_ho_deltas = Matrix.multiply(gradients, hidden_T);
// Adjust the weights by deltas
this.weights_ho.add(weight_ho_deltas);
// Adjust the bias by its deltas (which is just the gradients)
this.bias_o.add(gradients);
// Calculate the hidden layer errors
let who_t = Matrix.transpose(this.weights_ho);
let hidden_errors = Matrix.multiply(who_t, output_errors);
// Calculate hidden gradient
let hidden_gradient = Matrix.map(hidden, this.activation_function.dfunc);
hidden_gradient.multiply(hidden_errors);
hidden_gradient.multiply(this.learning_rate);
// Calcuate input->hidden deltas
let inputs_T = Matrix.transpose(inputs);
let weight_ih_deltas = Matrix.multiply(hidden_gradient, inputs_T);
this.weights_ih.add(weight_ih_deltas);
// Adjust the bias by its deltas (which is just the gradients)
this.bias_h.add(hidden_gradient);
// outputs.print();
// targets.print();
// error.print();
}
serialize() {
return JSON.stringify(this);
}
static deserialize(data) {
if (typeof data == 'string') {
data = JSON.parse(data);
}
let nn = new NeuralNetwork(data.input_nodes, data.hidden_nodes, data.output_nodes);
nn.weights_ih = Matrix.deserialize(data.weights_ih);
nn.weights_ho = Matrix.deserialize(data.weights_ho);
nn.bias_h = Matrix.deserialize(data.bias_h);
nn.bias_o = Matrix.deserialize(data.bias_o);
nn.learning_rate = data.learning_rate;
return nn;
}
// Adding function for neuro-evolution
copy() {
return new NeuralNetwork(this);
}
// Accept an arbitrary function for mutation
mutate(func) {
this.weights_ih.map(func);
this.weights_ho.map(func);
this.bias_h.map(func);
this.bias_o.map(func);
}
}
/*********************************************************************************************************************************************
I tried to write a multilayer neural network training model in nn.js. I spent a lot of time getting it to work,
but I failed, and I'm chagrined that I don't know what the problem is. It dose nnot work.
Reference websit :https://towardsdatascience.com/the-multi-channel-neural-network-26551bdfab6c
*************************************************************************************************************************************************/
/* Define the class of multilayer neural networks */
class NeuralNetworkMulti {
constructor(t, e) {
this.nodes = t, this.lr = e || .01, this.activation = NeuralNetworkMulti.rhlu, this.dactivation = NeuralNetworkMulti.drhlu, this.weights = [], this.biases = [];
for (let t = 0; t < this.nodes.length - 1; t++) this.weights.push(new Matrix(this.nodes[t + 1], this.nodes[t]).randomize());
for (let t = 1; t < this.nodes.length; t++) this.biases.push(new Matrix(this.nodes[t], 1).randomize())
}
/* Static function: derivative function of the activation function*/
static dtanh(t) {
return 1 / pow(Math.cosh(t), 2)
}
static sigmoid(t) {
return 1 / (1 + Math.exp(-t))
}
static dsigmoid(t) {
return t * (1 - t)
}
static rhlu(t) {
return t < 0 ? 0 : t
}
static drhlu(t) {
return t < 0 ? 0 : 1
}
/* Function: Input t, calculated by neural network, return the output */
predict(t) {
let e = Matrix.fromArray(t);
for (let t = 0; t < this.weights.length; t++)(e = Matrix.multiply(this.weights[t], e)).add(this.biases[t]), e.map(this.activation);
return e.toArray()
}
/* Function: Training of the neural network, t is the network input, e is the label ground truth that should correspond to input t. */ train(t, e) {
let i = Matrix.fromArray(e), // e is converted into matrix form and assigned to i
s = Matrix.fromArray(this.predict(t)),// input t, call the prediction function, convert the output value to matrix form, and assign the input to s
a = [],
n = Matrix.fromArray(t);
for (let t = 0; t < this.weights.length; t++)
{
a.push(n), // Put n in list a, list FIFO
(n = Matrix.multiply(this.weights[t], n)).add(this.biases[t]), // 权重weight x 输入n + 偏置bias, 输出结果重新赋值给n
n.map(this.activation); // Execute the activation function on n
}
/* find the loss function, calculate the error, back propagate to update the weight parameters, gradient */
let o = Matrix.subtract(i, s),// Find the difference between the network output s and the label i, understood as a loss function
r = Matrix.map(s, this.dactivation);
r.multiply(o), r.multiply(this.lr);
for (let t = a.length - 1; t >= 0; t--) {
let e = Matrix.multiply(r, Matrix.transpose(a[t]));
this.weights[t].add(e), this.biases[t].add(r), o = Matrix.multiply(Matrix.transpose(this.weights[t]), o), (r = Matrix.map(a[t], this.dactivation)).multiply(o), r.multiply(this.lr)
}
}
/* Function: Get network model model, return network configuration with network nodes, learning rate, activation function, derivatives of activation function, weights, bias */ getModel() {
let t = this,
/* */
e = {
nodes: t.nodes,
lr: t.lr,
activation: t.activation,
dactivation: t.dactivation,
weights: [],
biases: []
};
/* Define the size (rows and columns) of the weights (itself a matrix), the data content */
for (let i of t.weights) {
let t = {
rows: i.rows,
cols: i.cols,
data: []
};
for (let e of i.data) {
let i = [];
for (let t of e) i.push(t);
t.data.push(i)
}
e.weights.push(t)
}
/* Define the size row and content of bias bias */
for (let i of t.biases) {
let t = {
rows: i.rows,
cols: i.cols,
data: i.data
};
e.biases.push(t)
}
return e
}
/* Function: Pass in the network configuration parameters and return the network model Model class */
static formModel(t) {
let e = new NeuralNetworkMulti(t.nodes, t.lr);
e.nodes = t.nodes, e.lr = t.lr, e.activation = t.activation, e.dactivation = t.dactivation;
for (let i = 0; i < e.weights.length; i++) {
e.weights[i].rows = t.weights[i].rows, e.weights[i].cols = t.weights[i].cols;
for (let s = 0; s < t.weights[i].rows; s++)
for (let a = 0; a < t.weights[i].cols; a++) e.weights[i].data[s][a] = t.weights[i].data[s][a];
e.weights[i].rows = t.weights[i].rows
}
return e
}
/* Function: Define model, set model */
copy() {
let t = this.getModel();
return NeuralNetworkMulti.formModel(t)
}
/* Perform the operation of function t on weights, biases, */
mutate(t) {
for (let e of this.weights)
e.map(t);
for (let e of this.biases)
e.map(t)
}
/* Function: Do merge operation on current network weight bias, merge content from t */
merge(t, e = .5) {
let i = 1 - e,
s = e;
for (let e = 0; e < this.nodes.length; e++)
if (this.nodes[e] != t.nodes[e]) return void console.error("Neural Networks can not be merged");
this.lr = this.lr * i + t.lr * s;
for (let e = 0; e < this.weights.length; e++)
for (let a = 0; a < this.weights[e].rows; a++)
for (let n = 0; n < this.weights[e].cols; n++)
this.weights[e].data[a][n] = this.weights[e].data[a][n] * i + t.weights[e].data[a][n] * s;
for (let e = 0; e < this.biases.length; e++)
for (let a = 0; a < this.biases[e].rows; a++)
for (let n = 0; n < this.biases[e].cols; n++)
this.biases[e].data[a][n] = this.biases[e].data[a][n] * i + t.biases[e].data[a][n] * s;
return this
}
/* Set the activation function and the derivative of the activation function */
setActivation(t, e) {
this.activation = t, this.dactivation = e
}
/* set the learning rate */
setLearningRate(t) {
this.lr = t
}
}
//************************************************************* nn.js code end ******************************************************************************
// --- defined by MNIST - do not change these ---------------------------------------
const PIXELS = 28; // images in data set are tiny
const PIXELSSQUARED = PIXELS * PIXELS;
// number of training and test exemplars in the data set:
const NOTRAIN = 60000;
const NOTEST = 10000;
//--- can modify all these --------------------------------------------------
// no of nodes in network
const noinput = PIXELSSQUARED;
//const nohidden = 128;
const nohidden = 24; //change the hidden number 60 to 24
const nooutput = 10;
const learningrate = 0.1; // default 0.1
// should we train every timestep or not
let do_training = true;
// how many to train and test per timestep
const TRAINPERSTEP = 30;
const TESTPERSTEP = 5;
// multiply it by this to magnify for display
const ZOOMFACTOR = 7;
const ZOOMPIXELS = ZOOMFACTOR * PIXELS;
// 3 rows of
// large image + 50 gap + small image
// 50 gap between rows
//const canvaswidth = ( PIXELS + ZOOMPIXELS ) + 50;
//const canvasheight = ( ZOOMPIXELS * 3 ) + 100;
//**********************************************
const canvaswidth = 2 * ZOOMPIXELS + 120;
const canvasheight = 3 * ZOOMPIXELS + 102;
const doodlewidth = PIXELS;
const DOODLE_THICK = 15;
///////*****************************
//const DOODLE_THICK = 18; // thickness of doodle lines
const DOODLE_BLUR = 3; // blur factor applied to doodles
let mnist;
// all data is loaded into this
// mnist.train_images
// mnist.train_labels
// mnist.test_images
// mnist.test_labels
let nn, canvas, dst, src, hierarchy, contours, img, diffX, diffY, M, cvOutput = [];
let trainrun = 1;
let train_index = 0;
let testrun = 1;
let test_index = 0;
let total_tests = 0;
let total_correct = 0;
// images in LHS:
let doodle, demo;
let doodle_exists = false;
let demo_exists = false;
let mousedrag = false; // are we in the middle of a mouse drag drawing?
let hidden_no_slider, learning_rate_slider;
// save inputs to global var to inspect
// type these names in console
var train_inputs, test_inputs, demo_inputs, doodle_inputs;
// Matrix.randomize() is changed to point to this. Must be defined by user of Matrix.
function randomWeight()
{
return ( AB.randomFloatAtoB ( -0.5, 0.5 ) );
// Coding Train default is -1 to 1
}
// CSS trick
// make run header bigger
$("#runheaderbox").css ( { "max-height": "95vh" } );
//--- start of AB.msgs structure: ---------------------------------------------------------
// We output a serious of AB.msgs to put data at various places in the run header
var thehtml;
// 1 Doodle header
thehtml = "<hr> <h1> 1. Doodle </h1> Top row: Doodle (left) and shrunk (right). <br> " +
" Draw your doodle in top LHS. <button onclick='wipeDoodle();' class='normbutton' >Clear doodle</button> <br> ";
AB.msg ( thehtml, 1 );
// 2 Doodle variable data (guess)
// 3 Training header
thehtml = "<hr> <h1> 2. Training </h1> Middle row: Training image magnified (left) and original (right). <br> " +
" <button onclick='do_training = false;' class='normbutton' >Stop training</button> <br> ";
AB.msg ( thehtml, 3 );
// 4 variable training data
// 5 Testing header
thehtml = "<h3> Hidden tests </h3> " ;
AB.msg ( thehtml, 5 );
// 6 variable testing data
// 7 Demo header
thehtml = "<hr> <h1> 3. Demo </h1> Bottom row: Test image magnified (left) and original (right). <br>" +
" The network is <i>not</i> trained on any of these images. <br> " +
" <button onclick='makeDemo();' class='normbutton' >Demo test image</button> <br> ";
AB.msg ( thehtml, 7 );
// 8 Demo variable data (random demo ID)
// 9 Demo variable data (changing guess)
const greenspan = "<span style='font-weight:bold; font-size:x-large; color:darkgreen'> " ;
//--- end of AB.msgs structure: ---------------------------------------------------------
// function setup()
// {
// createCanvas ( canvaswidth, canvasheight );
// doodle = createGraphics ( ZOOMPIXELS, ZOOMPIXELS ); // doodle on larger canvas
// doodle.pixelDensity(1);
// // JS load other JS
// // maybe have a loading screen while loading the JS and the data set
// AB.loadingScreen();
// $.getScript ( "/uploads/codingtrain/matrix.js", function()
// { //https://ancientbrain.com/uploads/finally/nn.js
// //$.getScript ( "/uploads/finally/nn.js", function()
// $.getScript ( "/uploads/codingtrain/mnist.js", function()
// {
// console.log ("All JS loaded");
// nn = new NeuralNetwork( noinput, nohidden, nooutput );
// nn.setLearningRate ( learningrate );
// loadData();
// });
// });
// }
//*******************************************************Yuwei's code: re-write the setsup function *************************************************************
function setup() {
//define the canvas
canvas = createCanvas(canvaswidth, canvasheight);
canvas.position(10, 20);
canvas.background(51, 51, 51);
// doodle on larger canvas
doodle = createGraphics(ZOOMPIXELS, ZOOMPIXELS);
doodle.pixelDensity(1);
wipeDoodle();
//Draw a rectangle s the Doodle area
strokeWeight(3);
stroke(255,0,0);
rect(0, 0, ZOOMPIXELS, ZOOMPIXELS);
textSize(20);
textAlign(CENTER),
text("DOODLE AREA", ZOOMPIXELS / 2, ZOOMPIXELS / 2.2);
//image train area
rect(0, ZOOMPIXELS + 50, ZOOMPIXELS, ZOOMPIXELS);
//demo area
rect(0, canvasheight - ZOOMPIXELS - 2, ZOOMPIXELS, ZOOMPIXELS);
textSize(16);
textAlign(CENTER);
text(" DEMO AREA ", 100, canvasheight - ZOOMPIXELS / 1.8);
//rectangle for slider
rect(ZOOMPIXELS + 20, ZOOMPIXELS + 100, ZOOMPIXELS + 90, ZOOMPIXELS + 100);
text("HIDDEN_NO SLIDER ", ZOOMPIXELS + 120, ZOOMPIXELS + 130);
hidden_no_slider = createSlider(10, 300, nohidden);
hidden_no_slider.position(ZOOMPIXELS + 100, ZOOMPIXELS + 180);
//learning rate slider
text("LEARNING_RATE SLIDER ", ZOOMPIXELS + 120, ZOOMPIXELS + 230);
learning_rate_slider = createSlider(0.01, 0.2, learningrate);
learning_rate_slider.position(ZOOMPIXELS + 100, ZOOMPIXELS + 280);
// console.log(hidden_no_slider.value);
// // JS load other JS
// // maybe have a loading screen while loading the JS and the data set
AB.loadingScreen();
$.getScript("/uploads/finally/opencv.js", function() { //here I import the opencv API to deal with the doodle picture
$.getScript("/uploads/codingtrain/matrix.js", function() {
$.getScript("/uploads/finally/nn.js", function() {
$.getScript("/uploads/codingtrain/mnist.js", function() {
console.log("All JS loaded");
(nn = new NeuralNetwork(noinput, nohidden, nooutput)).setLearningRate(learningrate);
loadData();
})
})
})
})
}
//***************************************************************************************setup function end ************************************************88
// load data set from local file (on this server)
function loadData()
{
loadMNIST ( function(data)
{
mnist = data;
console.log ("All data loaded into mnist object:")
console.log(mnist);
AB.removeLoading(); // if no loading screen exists, this does nothing
});
}
function getImage ( img ) // make a P5 image object from a raw data array
{
let theimage = createImage (PIXELS, PIXELS); // make blank image, then populate it
theimage.loadPixels();
for (let i = 0; i < PIXELSSQUARED ; i++)
{
let bright = img[i];
let index = i * 4;
theimage.pixels[index + 0] = bright;
theimage.pixels[index + 1] = bright;
theimage.pixels[index + 2] = bright;
theimage.pixels[index + 3] = 255;
}
theimage.updatePixels();
return theimage;
}
function getInputs ( img ) // convert img array into normalised input array
{
let inputs = [];
for (let i = 0; i < PIXELSSQUARED ; i++)
{
let bright = img[i];
inputs[i] = bright / 255; // normalise to 0 to 1
}
return ( inputs );
}
function trainit (show) // train the network with a single exemplar, from global var "train_index", show visual on or off
{
let img = mnist.train_images[train_index];
let label = mnist.train_labels[train_index];
// optional - show visual of the image
if (show)
{
var theimage = getImage ( img ); // get image from data array
image ( theimage, 0, ZOOMPIXELS+50, ZOOMPIXELS, ZOOMPIXELS ); // magnified
image ( theimage, ZOOMPIXELS+50, ZOOMPIXELS+50, PIXELS, PIXELS ); // original
}
// set up the inputs
let inputs = getInputs ( img ); // get inputs from data array
// set up the outputs
let targets = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0];
targets[label] = 1; // change one output location to 1, the rest stay at 0
// console.log(train_index);
// console.log(inputs);
// console.log(targets);
train_inputs = inputs; // can inspect in console
nn.train ( inputs, targets );
thehtml = " trainrun: " + trainrun + "<br> no: " + train_index ;
AB.msg ( thehtml, 4 );
train_index++;
if ( train_index == NOTRAIN )
{
train_index = 0;
console.log( "finished trainrun: " + trainrun );
trainrun++;
}
}
function testit() // test the network with a single exemplar, from global var "test_index"
{
let img = mnist.test_images[test_index];
let label = mnist.test_labels[test_index];
// set up the inputs
let inputs = getInputs ( img );
test_inputs = inputs; // can inspect in console
let prediction = nn.predict(inputs); // array of outputs
let guess = findMax(prediction); // the top output
total_tests++;
if (guess == label) total_correct++;
let percent = (total_correct / total_tests) * 100 ;
thehtml = " testrun: " + testrun + "<br> no: " + total_tests + " <br> " +
" correct: " + total_correct + "<br>" +
" score: " + greenspan + percent.toFixed(2) + "</span>";
AB.msg ( thehtml, 6 );
test_index++;
if ( test_index == NOTEST )
{
console.log( "finished testrun: " + testrun + " score: " + percent.toFixed(2) );
testrun++;
test_index = 0;
total_tests = 0;
total_correct = 0;
}
}
//--- find no.1 (and maybe no.2) output nodes ---------------------------------------
// (restriction) assumes array values start at 0 (which is true for output nodes)
//****************************************** change the find12 function to output the no.1 , no.1 and no.3 nodes***********************************************
function find123(a) // return array showing indexes of no.1 and no.2 values in array
{
let no1 = 0;
let no2 = 0;
let no1value = 0;
let no2value = 0;
let no3 = 0;
let no3value = 0;
for (let i = 0; i < a.length; i++)
{
a[i] > no1value ? (no1 = i, no1value = a[i]) :
a[i] > no2value ? (no2 = i, no2value = a[i]) :
a[i] > no3value && (no3 = i, no3value = a[i]);
}
var b = [ no1, no2, no3];
return b;
}
//**************************************************************************find123 function end ******************************************************
// just get the maximum - separate function for speed - done many times
// find our guess - the max of the output nodes array
function findMax (a)
{
let no1 = 0;
let no1value = 0;
for (let i = 0; i < a.length; i++)
{
if (a[i] > no1value)
{
no1 = i;
no1value = a[i];
}
}
return no1;
}
// --- the draw function -------------------------------------------------------------
// every step:
// function draw()
// {
// // check if libraries and data loaded yet:
// if ( typeof mnist == 'undefined' ) return;
// // how can we get white doodle on black background on yellow canvas?
// // background('#ffffcc'); doodle.background('black');
// background ('black');
// if ( do_training )
// {
// // do some training per step
// for (let i = 0; i < TRAINPERSTEP; i++)
// {
// if (i == 0) trainit(true); // show only one per step - still flashes by
// else trainit(false);
// }
// // do some testing per step
// for (let i = 0; i < TESTPERSTEP; i++)
// testit();
// }
// // keep drawing demo and doodle images
// // and keep guessing - we will update our guess as time goes on
// if ( demo_exists )
// {
// drawDemo();
// guessDemo();
// }
// if ( doodle_exists )
// {
// drawDoodle();
// guessDoodle();
// }
// // detect doodle drawing
// // (restriction) the following assumes doodle starts at 0,0
// if ( mouseIsPressed ) // gets called when we click buttons, as well as if in doodle corner
// {
// // console.log ( mouseX + " " + mouseY + " " + pmouseX + " " + pmouseY );
// var MAX = ZOOMPIXELS + 20; // can draw up to this pixels in corner
// if ( (mouseX < MAX) && (mouseY < MAX) && (pmouseX < MAX) && (pmouseY < MAX) )
// {
// mousedrag = true; // start a mouse drag
// doodle_exists = true;
// doodle.stroke('white');
// doodle.strokeWeight( DOODLE_THICK );
// doodle.line(mouseX, mouseY, pmouseX, pmouseY);
// }
// }
// else
// {
// // are we exiting a drawing
// if ( mousedrag )
// {
// mousedrag = false;
// // console.log ("Exiting draw. Now blurring.");
// doodle.filter (BLUR, DOODLE_BLUR); // just blur once
// // console.log (doodle);
// }
// }
// }
//**********************************************************Yuwei's code, re-write the draw()function *****************************************************************************8
function draw() {
if (void 0 !== mnist)
{
if (strokeWeight(3),
stroke(255,0,0),
rect(0, 0, ZOOMPIXELS, ZOOMPIXELS),
textSize(20),
textAlign(CENTER),
text("DOODLE AREA", ZOOMPIXELS / 2, ZOOMPIXELS / 2.2), do_training)
// do some training per step
{
for (let t = 0; t < TRAINPERSTEP; t++) trainit(0 === t); // show only one per step - still flashes by
for (let t = 0; t < TESTPERSTEP; t++) testit() // do some testing per step
}
// detect doodle drawing
// (restriction) the following assumes doodle starts at 0,0
if (demo_exists && (drawDemo(),
guessDemo()), doodle_exists && (drawDoodle(),
guessDoodle()), mouseIsPressed)
{
var MAX = ZOOMPIXELS - 2;
mouseX < MAX && mouseY < MAX && pmouseX < MAX && pmouseY < MAX &&
(mousedrag = !0, doodle_exists = !0, doodle.stroke("cream"),
strokeJoin(BEVEL), doodle.strokeWeight(DOODLE_THICK),
doodle.line(mouseX, mouseY, pmouseX, pmouseY))
} else if (mousedrag) {
mousedrag = !1,
console.log("Doodle detected"),
(img = doodle.get()).resize(PIXELS, PIXELS),
img.loadPixels(),
imagedata = img.imageData,
src = cv.matFromImageData(imagedata), //create a cv.Mat
dst = cv.Mat.zeros(src.cols, src.rows, cv.CV_8UC3), //creating a black diagram with 0 for each channel of each pixel,Scalar(0,0,0);
cv.cvtColor(src, src, cv.COLOR_RGBA2GRAY, 0), //Conversion from RBG color space to grayscale space
cv.threshold(src, src, 120, 255, cv.THRESH_BINARY), //Image denoising to better snap out numbers
contours = new cv.MatVector, //create a new cv.MatVector and a new cv.Mat
hierarchy = new cv.Mat,
cv.findContours(src, contours, hierarchy, cv.RETR_CCOMP, cv.CHAIN_APPROX_SIMPLE); // find the contour of the doodle digit
let contour = contours.get(0), //get contours
m = cv.moments(contour, 0); //calculate moments of binary image
//# calculate x,y coordinate of center
M = m.m00;
let x = Math.round(m.m10 / m.m00 * 100) / 100,
y = Math.round(m.m01 / m.m00 * 100) / 100,
xcen = PIXELS / 2,
ycen = PIXELS / 2;
diffX = Math.round(xcen - x),
diffY = Math.round(ycen - y);
let r = cv.matFromArray(2, 3, cv.CV_64FC1, [1, 0, diffX, 0, 1, diffY]); //a new cv.Mat
return dsize = new cv.Size(src.rows, src.cols), //reset image size,Zoom in on the digit
//Imitation of the image using the transformation matrix dst
cv.warpAffine(src, dst, r, dsize, cv.INTER_LINEAR,
cv.BORDER_CONSTANT, new cv.Scalar),
//output final doodle digit picture
cvOutput = getInputs(dst.data8S),
image(img, ZOOMPIXELS + 120 + diffX * ZOOMFACTOR, 0 + diffY * ZOOMFACTOR, ZOOMPIXELS, ZOOMPIXELS),
cvOutput
}
}
}
//**********************************************************draw() function end*****************************************************************************8
//--- demo -------------------------------------------------------------
// demo some test image and predict it
// get it from test set so have not used it in training
function makeDemo()
{
demo_exists = true;
var i = AB.randomIntAtoB ( 0, NOTEST - 1 );
demo = mnist.test_images[i];
var label = mnist.test_labels[i];
thehtml = "Test image no: " + i + "<br>" +
"Classification: " + label + "<br>" ;
AB.msg ( thehtml, 8 );
// type "demo" in console to see raw data
}
function drawDemo()
{
var theimage = getImage ( demo );
// console.log (theimage);
image ( theimage, 0, canvasheight - ZOOMPIXELS, ZOOMPIXELS, ZOOMPIXELS ); // magnified
image ( theimage, ZOOMPIXELS+50, canvasheight - ZOOMPIXELS, PIXELS, PIXELS ); // original
}
function guessDemo()
{
let inputs = getInputs ( demo );
demo_inputs = inputs; // can inspect in console
let prediction = nn.predict(inputs); // array of outputs
let guess = findMax(prediction); // the top output
thehtml = " We classify it as: " + greenspan + guess + "</span>" ;
AB.msg ( thehtml, 9 );
}
//--- doodle -------------------------------------------------------------
function drawDoodle()
{
// doodle is createGraphics not createImage
let theimage = doodle.get();
// console.log (theimage);
image ( theimage, 0, 0, ZOOMPIXELS, ZOOMPIXELS ); // original
image ( theimage, ZOOMPIXELS+50, 0, PIXELS, PIXELS ); // shrunk
}
let lastT = 0
function guessDoodle() {
(img = doodle.get()).resize(PIXELS, PIXELS), img.loadPixels();
let t = [];
for (let e = 0; e < PIXELSSQUARED; e++) t[e] = img.pixels[4 * e] / 255;
let e = Array.from(t);
for (let t = 0; t < e.length; t++) e[t] = cvOutput[t];
t = Array.from(e);
for (let e = 0; e < t.length; e++) t[e] = -255 * t[e];
doodle_inputs = t;
let n = find123(nn.predict(t));
thehtml = " We classify it as: " + greenspan + n[0] + "</span> <br> No.2 guess is: " + greenspan + n[1] + "</span> <br>No.3 guess is: " + greenspan + n[2] + "</span>", AB.msg(thehtml, 2), AB.msg(thehtml, 2)
}
function wipeDoodle()
{
doodle_exists = false;
doodle.background('black');
}
// --- debugging --------------------------------------------------
// in console
// showInputs(demo_inputs);
// showInputs(doodle_inputs);
function showInputs ( inputs )
// display inputs row by row, corresponding to square of pixels
{
var str = "";
for (let i = 0; i < inputs.length; i++)
{
if ( i % PIXELS == 0 ) str = str + "\n"; // new line for each row of pixels
var value = inputs[i];
str = str + " " + value.toFixed(2) ;
}
console.log (str);
}