// Cloned by Thomas Mc Cann on 4 Dec 2019 from World "Character recognition neural network" by "Coding Train" project
// Please leave this clone trail here.
// Port of Character recognition neural network from here:
// https://github.com/CodingTrain/Toy-Neural-Network-JS/tree/master/examples/mnist
// with many modifications
// --- 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;
// **************************** THOMAS MC CANN ****************************
//
//
//
// **************************** THOMAS MC CANN ****************************
const nohidden = 32;
const nooutput = 10;
// **************************** THOMAS MC CANN ****************************
//
// learningrate = 0.1 unchanged after experimentation
//
// **************************** THOMAS MC CANN ****************************
let 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;
// **************************** THOMAS MC CANN ****************************
//
// DOODLE_THICK & DOODLE_BLUR slightly changed
//
// **************************** THOMAS MC CANN ****************************
const DOODLE_THICK = 17; // 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;
let trainrun = 1;
let train_index = 0;
// **************************** THOMAS MC CANN ****************************
//
// Adding a training order array that will be shuffled on each
// training run after the first training run is complete.
// It will be initiliased in setup() in normal ascending order and the
// order shuffled at the end of each training run.
// This should reduce the chance of over fit occuring if the NN
// repeatedly nudges it weights in the same order on each run.
//
// **************************** THOMAS MC CANN ****************************
let trainingOrder = [];
let testrun = 1;
let test_index = 0;
// Holds the number of correct tests and total tests for all runs
let total_tests = 0;
let total_correct = 0;
// **************************** THOMAS MC CANN ****************************
//
// Variable to hold the overall performance of the network over all runs
// Some variable names have ben mofiied for measuring performance over
// one/current training run.
//
// **************************** THOMAS MC CANN ****************************
let overallPerformance = 0;
// Holds the number of correct tests and total tests for a given run
let correctOnThisRun = 0;
let totalTestsOnthisRun = 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?
// 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: ---------------------------------------------------------
// **************************** THOMAS MC CANN ****************************
//
// Initialise the traing order array to a default ascending triang order
// 0 to NOTRAIN -1
//
// **************************** THOMAS MC CANN ****************************
function initialiseTrainingOrder(size){
for (let i = 0; i < size ; i++)
{
trainingOrder[i] = i;
}
}
// **************************** THOMAS MC CANN ****************************
//
// Retreive the next training order from the
// trainingOrder array
//
// **************************** THOMAS MC CANN ****************************
function getNextTrainingIndex(index){
return trainingOrder[index];
}
// **************************** THOMAS MC CANN ****************************
//
// Shuufel the training order called at the end of a trining run
// Does not affect the MMIST data structures.
//
// **************************** THOMAS MC CANN ****************************
function shuffleTrainingOrder(){
shuffle(trainingOrder, true);
}
// **************************** THOMAS MC CANN ****************************
//
// Calculate current performance for stats and adjustinng the learning rate
//
// **************************** THOMAS MC CANN ****************************
function currentRunPerformance(){
return ((correctOnThisRun / totalTestsOnthisRun) * 100) ;
}
// **************************** THOMAS MC CANN ****************************
//
// Calculate overall performance for stats and adjustinng the learning rate
//
// **************************** THOMAS MC CANN ****************************
function calcOverallPerfromance(){
return ( (total_correct / total_tests) * 100);
}
// **************************** THOMAS MC CANN ****************************
//
// Adjust learning rate based on current and overall performance
//
// **************************** THOMAS MC CANN ****************************
function adjustLearningRate(currentPerformance, overAllPerformance){
var current = currentPerformance;
var overall = overAllPerformance;
// console.log(overAllPerformance);
if(current > overall){
// Increase learning rate
learningrate = learningrate + ((overall-current)/(overall));
// console.log((overall-current)/(overall));
if(Number.isNaN(learningrate)){
learningrate = 1.0001;
}
}
else{
// Decrease learning rate
learningrate = learningrate - ((current-overall)/(overall));
// console.log((current-overall)/(overall));
if(Number.isNaN(learningrate)){
learningrate = .99999;
}
}
}
// **************************** THOMAS MC CANN ****************************
//
// softMax activation function. Tried with the NN.js code
//
// **************************** THOMAS MC CANN ****************************
function softmax(arr) {
return arr.map(function(value,index) {
return Math.exp(value) / arr.map( function(y /*value*/){ return Math.exp(y) } ).reduce( function(a,b){ return a+b })
})
}
function setup()
{
// **************************** THOMAS MC CANN ****************************
//
// Initialise the trainingOrder array from 0 to NOTRIAN -1
//
// **************************** THOMAS MC CANN ****************************
initialiseTrainingOrder(NOTRAIN);
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()
{
$.getScript ( "/uploads/codingtrain/nn.js", function()
{
$.getScript ( "/uploads/codingtrain/mnist.js", function()
{
console.log ("All JS loaded");
nn = new NeuralNetwork( noinput, nohidden, nooutput );
nn.setLearningRate ( learningrate );
loadData();
});
});
});
}
// 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);
console.log(mnist.train_images);
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[getNextTrainingIndex(train_index)];
let label = mnist.train_labels[getNextTrainingIndex(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 );
// **************************** THOMAS MC CANN ****************************
//
// Shuffle the trainingOrder so that the NN does not
// get retrained in the same order.
//
// **************************** THOMAS MC CANN ****************************
shuffleTrainingOrder();
console.log("Training order has been shuffled to prevent over training.")
trainrun++;
}
}
// **************************** THOMAS MC CANN ****************************
//
// Used for stats
//
// **************************** THOMAS MC CANN ****************************
let bestPerformanceOnRun = 0;
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++;
totalTestsOnthisRun++
if (guess == label) {
total_correct++;
correctOnThisRun++;
}
// **************************** THOMAS MC CANN ****************************
//
// Calculate best performance - used in stats
//
// **************************** THOMAS MC CANN ****************************
if(currentRunPerformance() > bestPerformanceOnRun){
bestPerformanceOnRun = currentRunPerformance();
}
thehtml = " Testrun: " + testrun + "<br> no: " + total_tests + " <br> " +
" Correct: " + total_correct + "<br>" +
" Score: " + greenspan + currentRunPerformance().toFixed(2) + "</span>" + "<br>"
+ "Best of run : " + bestPerformanceOnRun.toFixed(2) + "%" + "<br>"
+ "Overall Performance : " + calcOverallPerfromance().toFixed(2) + "%" + "<br>"
+ "Learning Rate : " + learningrate.toFixed(10);
AB.msg ( thehtml, 6 );
test_index++;
if ( test_index == NOTEST )
{
console.log( "finished testrun: " + testrun + " score: " + currentRunPerformance() );
testrun++;
test_index = 0;
// **************************** THOMAS MC CANN ****************************
//
// Reset some stat variable at the end of each run
//
// **************************** THOMAS MC CANN ****************************
correctOnThisRun = 0;
totalTestsOnthisRun = 0;
bestPerformanceOnRun = 0;
}
// **************************** THOMAS MC CANN ****************************
//
// Adjust learning rate based on current and overall performance
// Ending up getting a NaN in adjustLearningRate() which I can't figure out
// as all variables are numbers ad initialised
//
// **************************** THOMAS MC CANN ****************************
adjustLearningRate(currentRunPerformance(), calcOverallPerfromance());
}
//--- find no.1 (and maybe no.2) output nodes ---------------------------------------
// (restriction) assumes array values start at 0 (which is true for output nodes)
function find12 (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;
for (let i = 0; i < a.length; i++)
{
if (a[i] > no1value)
{
no1 = i;
no1value = a[i];
}
else if (a[i] > no2value)
{
no2 = i;
no2value = a[i];
}
}
var b = [ no1, no2 ];
return b;
}
// 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);
}
}
}
//--- 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
}
function guessDoodle()
{
// doodle is createGraphics not createImage
let img = doodle.get();
// **************************** THOMAS MC CANN ****************************
//
// The basic idea that I think would be effective is to locate the rectangular
// boundary of the bright area of the users doodle inside the 28x28 image.
// If we think of the (unprocessed) user doodle image as a matrix or 2D array
// then the boundaries can be located by finding the highest, lowest, left most
// and right most bright pixels on each row/column.
// This ‘bright’ area/portion can then then be copied, resized, scaled and
// centered within the center 20x20 pixels of a new 28x28 pixel image.
// This new image which still represents the users doodle would lead to a
// much higher prediction accuracy as it would be more standardised and more
// closely conform and resemble the MNIST training datasets which the NN
// has trained on.
//
// **************************** THOMAS MC CANN ****************************
let preProcessedImage = createImage(PIXELS, PIXELS);
// console.log(img); // 196 pixels
img.resize ( PIXELS, PIXELS );
// **************************** THOMAS MC CANN ****************************
//
// user doodle image processing to scale, resize and center would follow
// from here but not finalised.
//
// **************************** THOMAS MC CANN ****************************
getDoodleImageBoundaries(img);
// preProcessedImage = centerScaleResizeImage(img)
img.loadPixels();
// set up inputs
let inputs = [];
for (let i = 0; i < PIXELSSQUARED ; i++)
{
inputs[i] = img.pixels[i * 4] / 255;
}
doodle_inputs = inputs; // can inspect in console
// feed forward to make prediction
let prediction = nn.predict(inputs); // array of outputs
let b = find12(prediction); // get no.1 and no.2 guesses
thehtml = " We classify it as: " + greenspan + b[0] + "</span> <br>" +
" No.2 guess is: " + greenspan + b[1] + "</span>";
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);
}
// **************************** THOMAS MC CANN ****************************
//
// getDoodleImageBoundaries(img) Takes a user doodle image and finds the
// rectangular boundaries of the area ofthe image in order to pocress it.
// I was unable to get it working fully.
//
// **************************** THOMAS MC CANN ****************************
function getDoodleImageBoundaries(img){
let theimage = img.get();
// theimage has 28*28*4 pixels (r,g,b,a)
theimage.loadPixels();
let ImageMostLeftNonBlankPixel = 27;
let RowMostleftNonBlankPixel = 0;
let RowMostRightNonBlankPixel = 27
let ImageMostRightNonBlankPixel = 27;
let pixelIindex = 0;
// Get left row index where doodle starts
for (let row = 0; row < PIXELS ; row++)
{
for (let col = 0; col < PIXELS; col++)
{
pixelIndex = (row+col*PIXELS)*4;
// Add the values of the first 3 (of 4 ) pixels
var rgbPixelValue = theimage.pixels[pixelIndex] + theimage.pixels[pixelIndex + 1] + theimage.pixels[pixelIndex + 2];
if(rgbPixelValue > 0){
RowMostleftNonBlankPixel = row;
if(RowMostleftNonBlankPixel < ImageMostLeftNonBlankPixel){
ImageMostLeftNonBlankPixel = RowMostleftNonBlankPixel;
break;
}
}
}
}
// Get left row index where doodle starts
for (let row = PIXELS-1; row >0 ; row--)
{
// Get right row index where doodle ends
for (let col = 0; col < PIXELS; col++)
{
pixelIndex = (row+col*PIXELS)*4;
// Add the values of the first 3 (of 4 ) pixels
var rgbPixelValue = theimage.pixels[pixelIndex] + theimage.pixels[pixelIndex + 1] + theimage.pixels[pixelIndex + 2];
if(rgbPixelValue > 0){
RowMostRightNonBlankPixel = row;
if(RowMostRightNonBlankPixel < ImageMostRightNonBlankPixel){
ImageMostRightNonBlankPixel = RowMostRightNonBlankPixel;
break;
}
}
}
}
console.log("Row left most non blank pixel: " + ImageMostLeftNonBlankPixel);
console.log("Row right most non blank pixel: " + ImageMostRightNonBlankPixel);
theimage.updatePixels();
}
// **************************** THOMAS MC CANN ****************************
//
// Resize, scale and center in the centre 20x20 pixels of a new 28x28 image
// Which would then be used as input to guess the number drawn.
// Only a stub.
//
// **************************** THOMAS MC CANN ****************************
function centerScaleResizeImage(img){
return img;
}