/*Student Name: Meenu Mathew
Student Number: 20211335
Assignment 2: Character Recognition -Neural Network*/
// --- 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 = 64;
const nooutput = 10;
const learningrate = 0.1; // default 0.1 code by meenu
// should we train every timestep or not
let do_training = true;
let doodle_exists = true;// code by meenu //to load doodle canvas when the screen loaded
// 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;
const canvaswidth = ( PIXELS + ZOOMPIXELS )+ 100;
//const canvasheight = ( PIXELS + ZOOMPIXELS ) + 110;
const canvasheight = ( ZOOMPIXELS * 2)+ 100 ;
const DOODLE_THICK = 18; // thickness of doodle lines
const DOODLE_BLUR = 3; // blur factor applied to doodles
let mnist,nn, conv_net, conv_trainer, doodle;//code by Meenu
let trainrun = 1;
let train_index = 0;
let testrun = 1;
let test_index = 0;
let total_tests = 0;
let ts_total_correct = 0;
let conv_total_tests = 0;//code by Meenu
let conv_ts_total_correct = 0;//code by Meenu
// save inputs to global var to inspect
// type these names in console
var train_inputs, test_inputs, doodle_inputs;
// images in LHS:
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, 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
}
// make run header bigger
AB.headerCSS ( { "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
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);
doodle.background("black");
// 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();
});
});
});
//code by Meenu
// import convnetjs library
//The scrips are using and are based on Andrej Karpathy's ConvnetJS
$.getScript ( "/uploads/meenumathew/convnet-min.js", function()
{
conv_layer_defs = [];
conv_layer_defs.push({type:'input', out_sx:28, out_sy:28, out_depth:1});
conv_layer_defs.push({type:'conv', sx:3, filters:4, stride:1, pad:2, activation:'relu'});
conv_layer_defs.push({type:'pool', sx:3, stride:1});
conv_layer_defs.push({type:'conv', sx:3, filters:4, stride:1, pad:2, activation:'relu'});
conv_layer_defs.push({type:'pool', sx:3, stride:1});
/* conv_layer_defs.push({type:'conv', sx:3, filters:16, stride:1, pad:2, activation:'relu'});
conv_layer_defs.push({type:'pool', sx:3, stride:1});
conv_layer_defs.push({type:'conv', sx:3, filters:16, stride:1, pad:2, activation:'relu'});
conv_layer_defs.push({type:'pool', sx:3, stride:1});
conv_layer_defs.push({type:'conv', sx:3, filters:16, stride:1, pad:2, activation:'relu'});
conv_layer_defs.push({type:'pool', sx:3, stride:1});*/
conv_layer_defs.push({type:'softmax', num_classes:10});
conv_net = new convnetjs.Net();
conv_net.makeLayers(conv_layer_defs);
conv_trainer = new convnetjs.SGDTrainer(conv_net, {method:'adadelta', batch_size:12, l2_decay:0.01});
});
}
// 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;
}
//code by Meenu
// convert img array suitable for convoultional network
function getConvInputs(img) {
let conv_img = new convnetjs.Vol(PIXELS, PIXELS, 1, 0);
//console.log(conv_img);
for (let i = 0; i < PIXELSSQUARED ; i++) {
conv_img.w[i] = img[i];
}
return conv_img;
}
function getInputs( img ) // convert img array into normalised input array
{
let inputs = [];
for (let i = 0; i < PIXELSSQUARED ; i++)
{
inputs[i] = img[i]/ 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 tr_img = mnist.train_images[train_index];
let tr_label = mnist.train_labels[train_index];
// optional - show visual of the image
if (show)
{
var theimage = getImage(tr_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 tr_inputs = getInputs(tr_img); // get inputs from data array
let tr_conv_inputs = getConvInputs(tr_inputs); // get inputs for convulational network
let tr_targets = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0]; // set up the outputs
tr_targets[tr_label] = 1; // change one output location to 1, the rest stay at 0
nn.train ( tr_inputs, tr_targets );
conv_trainer.train(tr_conv_inputs,tr_label);//code by Meenu
thehtml = " Train run: " + trainrun + "<br> no: " + train_index ;
AB.msg ( thehtml, 4 );
train_index++;
if (train_index == NOTRAIN )
{
train_index = 0;
console.log( "Finished Train run: " + trainrun );
trainrun++;
}
}
function testit() // test the network with a single exemplar, from global var "test_index"
{
let ts_img = mnist.test_images[test_index];
let ts_label = mnist.test_labels[test_index];
// set up the inputs
let ts_inputs = getInputs ( ts_img );
//test_inputs = inputs; // can inspect in console
// finding neural network guess
let ts_nn_prediction = nn.predict(ts_inputs); // array of outputs
let ts_nn_guess = findMax(ts_nn_prediction); // the top output
//code by Meenu
// finding covolutional neural network guess based on Andrej Karpathy's ConvnetJS
let ts_conv_inputs = getConvInputs(ts_inputs);//inputs for conv nw
let ts_conv_prediction = conv_net.forward(ts_conv_inputs).w; // array of outputs
let ts_conv_guess = findMax(ts_conv_prediction); // the top output
total_tests++;// calculate no of test performed
if (ts_nn_guess == ts_label) ts_total_correct++;//to find the total no of correct inputs obtained by NN
let percent = (ts_total_correct / total_tests) * 100 ;//find percentage of correct predicted output
//code by Meenu
conv_total_tests++;
if (ts_conv_guess == ts_label) conv_ts_total_correct++;//to find the total no of correct inputs obtained by NN
let conv_percent = (conv_ts_total_correct / conv_total_tests) * 100 ;//find percentage of correct predicted output
//code by Meenu
thehtml = " Test run: " + testrun + "<br> No: " + total_tests +
" <br/> Neural N/w Correct: " + ts_total_correct + " Conv N/w Correct: " + conv_ts_total_correct +
" <br/> Neural N/w Score( %): " + greenspan + percent.toFixed(2) + "</span> Conv N/w Score(%): " + greenspan + conv_percent.toFixed(2) + "</span>";
AB.msg ( thehtml, 6 );
test_index++;
if ( test_index == NOTEST )
{
console.log( "Finished testrun: " + testrun + " Neural N/w Score: " + percent.toFixed(2) +"Conv N/w Score:" + conv_percent.toFixed(2));
testrun++;
test_index = 0;
total_tests = 0;
ts_total_correct = 0;
conv_ts_total_correct=0; //code by Meenu
}
}
//--- 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) // new no1
{
// old no1 becomes no2
no2 = no1;
no2value = no1value;
// now put in the new no1
no1 = i;
no1value = a[i];
}
else if (a[i] > no2value) // new no2
{
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;
if ( do_training )
{
// code by meenu
//set bg as red
background ('red');
// do some training per step
for (let j = 0; j < TRAINPERSTEP; j++)
{
if (j == 0) trainit(true); // show only one per step - still flashes by
else trainit(false);
}
// do some testing per step
for (let j = 0; j < TESTPERSTEP; j++)
testit();
}
// -------------------------doodle images ------------------------------------
if(doodle_exists){
drawDoodle();
//code by Meenu
line(0, ZOOMPIXELS+12, canvaswidth, ZOOMPIXELS+12); //draw horizonatal to show seperation of each canvas
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.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);
}
}
}
//--- doodle -------------------------------------------------------------
function drawDoodle()
{
// doodle is createGraphics not createImage
let doodle_image = doodle.get();
// console.log (doodle_image);
image ( doodle_image, 0, 0, ZOOMPIXELS, ZOOMPIXELS ); // original image
image ( doodle_image, ZOOMPIXELS + 50 , 0, PIXELS, PIXELS ); // shrunk
}
function guessDoodle()
{
// doodle is createGraphics not createImage
let img = doodle.get();
img.resize ( PIXELS, PIXELS );
img.loadPixels();//Loads the pixel data for the display window into the pixels[] array
// set up inputs
//normalising inputs array
let inputs = [];
for (let i = 0; i < PIXELSSQUARED ; i++)
{
inputs[i] = img.pixels[i * 4] / 255;
}
// finding neural network guess
let nn_prediction = nn.predict(inputs); // array of outputs
let nn_guess = find12(nn_prediction); // get no.1 and no.2 guesses
//code by Meenu
// finding covolutional neural network guess based on Andrej Karpathy's ConvnetJS
let conv_img = new convnetjs.Vol(PIXELS, PIXELS, 1, 0);
//console.log(conv_img);
for (let i = 0; i < PIXELSSQUARED ; i++) {
conv_img.w[i] = inputs[i];
}
let conv_inputs = conv_img;
let conv_prediction = conv_net.forward(conv_inputs).w; // array of outputs
let conv_guess = find12(conv_prediction); // get no.1 and no.2 guesses
thehtml = " Neural Network classify it as: " + greenspan + nn_guess[0] + "</span> <br>" +
" No.2 guess is: " + greenspan + nn_guess[1] + "</span>"+
" <br/>Convolutional Neural Network(based on Andrej Karpathy's ConvnetJS) classify it as: " + greenspan + conv_guess[0] + "</span> <br>" +
" No.2 guess is: " + greenspan + conv_guess[1] + "</span>";
AB.msg ( thehtml, 2 );
}
function wipeDoodle()
{
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);
}