File size: 2,987 Bytes
4559903 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 | #include "ExpressionClassifier.h"
using namespace cv;
/*
sigma describes the classification sharpness. A larger sigma means the
boundary between different expressions is more blurry. It won't change
the classification, but will give you probabilities that are smoother.
*/
ExpressionClassifier::ExpressionClassifier()
:sigma(10.0) {
}
void ExpressionClassifier::save(string directory) const {
ofDirectory dir(directory);
dir.create(true);
for(int i = 0; i < size(); i++) {
string filename = dir.path() + "/" + expressions[i].getDescription() + ".yml";
cout << "saving to " << filename << endl;
expressions[i].save(filename);
}
}
void ExpressionClassifier::load(string directory) {
ofDirectory dir(directory);
dir.listDir();
int n = dir.size();
expressions.resize(n);
for(int i = 0; i < n; i++) {
expressions[i].load(dir.getPath(i));
}
}
unsigned int ExpressionClassifier::classify(const ofxFaceTracker& tracker) {
Mat cur;
tracker.getObjectPointsMat().copyTo(cur);
norm(cur);
int n = size();
probability.resize(n);
if(n == 0) {
return 0;
}
vector<vector<double> > val(n);
double sum = 0;
for(int i = 0; i < n; i++){
int m = expressions[i].size();
for(int j = 0; j < m; j++){
double v = norm(cur, expressions[i].getExample(j));
double p = exp(-v * v / sigma);
val[i].push_back(p);
sum += p;
}
}
for(int i = 0; i < n; i++){
probability[i] = 0;
int m = expressions[i].size();
for(int j = 0; j < m; j++) {
probability[i] += val[i][j];
}
probability[i] /= sum;
}
return getPrimaryExpression();
}
unsigned int ExpressionClassifier::getPrimaryExpression() const {
int maxExpression = 0;
double maxProbability = 0;
for(int i = 0; i < probability.size(); i++) {
double cur = getProbability(i);
if(cur > maxProbability) {
maxExpression = i;
maxProbability = cur;
}
}
return maxExpression;
}
double ExpressionClassifier::getProbability(unsigned int i) const {
if(i < probability.size()) {
return probability[i];
} else {
return 0;
}
}
string ExpressionClassifier::getDescription(unsigned int i) const {
return expressions[i].getDescription();
}
Expression& ExpressionClassifier::getExpression(unsigned int i) {
return expressions[i];
}
void ExpressionClassifier::setSigma(double sigma) {
this->sigma = sigma;
}
double ExpressionClassifier::getSigma() const {
return sigma;
}
unsigned int ExpressionClassifier::size() const {
return expressions.size();
}
void ExpressionClassifier::addExpression(string description) {
if(description == "") {
description = ofToString(expressions.size());
}
expressions.push_back(Expression(description));
}
void ExpressionClassifier::addExpression(Expression& expression) {
expressions.push_back(expression);
}
void ExpressionClassifier::addSample(const ofxFaceTracker& tracker) {
if(size() == 0) {
addExpression();
}
expressions.back().addSample(tracker.getObjectPointsMat());
}
void ExpressionClassifier::reset() {
expressions.clear();
} |