# Read the training set
train_images_filenames, train_labels = io.load_training_set()
print('Loaded {} train images.'.format(len(train_images_filenames)))
# Feature extraction with sift
print('Obtaining sift features...')
D, L, _, _ = feature_extraction.parallel_sift(train_images_filenames,
train_labels,
n_jobs=N_JOBS)
print('Time spend: {:.2f} s'.format(time.time() - start))
temp = time.time()
# Train Linear SVM classifier
print('Training the SVM classifier...')
lin_svm, std_scaler, pca = classification.train_rbf_svm(
D, L, model_name='final_noprob_sift_all_svm')
print('Time spend: {:.2f} s'.format(time.time() - temp))
temp = time.time()
# Read the test set
test_images_filenames, test_labels = io.load_test_set()
print('Loaded {} test images.'.format(len(test_images_filenames)))
# Feature extraction with sift, prediction with SVM and aggregation to obtain final class
print('Predicting test data...')
result = joblib.Parallel(n_jobs=N_JOBS, backend='threading')(
joblib.delayed(parallel_testing)(test_image, test_label, lin_svm,
std_scaler, None)
for test_image, test_label in zip(test_images_filenames, test_labels))
correct_class = [i[0] for i in result]
import mlcv.feature_extraction as feature_extraction
import mlcv.input_output as io
from scripts import SESSION1
if __name__ == '__main__':
start = time.time()
# Read the training set
train_images_filenames, train_labels = io.load_training_set()
print('Loaded {} train images.'.format(len(train_images_filenames)))
# Feature extraction with sift
print('Obtaining sift features...')
D, L, _, _ = feature_extraction.parallel_sift(train_images_filenames,
train_labels)
print('Time spend: {:.2f} s'.format(time.time() - start))
temp = time.time()
# Train Linear SVM classifier
print('Training the SVM classifier...')
svm, std_scaler, pca = classification.train_rbf_svm(
D,
L,
model_name=SESSION1['model'],
save_scaler=SESSION1['scaler'],
save_pca=SESSION1['pca'])
print('Time spend: {:.2f} s'.format(time.time() - temp))
temp = time.time()
print('\nTOTAL TRAINING TIME: {:.2f} s'.format(time.time() - start))
start = time.time()
# Read the training set
train_images_filenames, train_labels = io.load_training_set()
print('Loaded {} train images.'.format(len(train_images_filenames)))
# Feature extraction with surf
print('Obtaining surf features...')
D, L, _ = feature_extraction.parallel_surf(train_images_filenames,
train_labels)
print('Time spend: {:.2f} s'.format(time.time() - start))
temp = time.time()
# Train Linear SVM classifier
print('Training the SVM classifier...')
lin_svm, std_scaler, pca = classification.train_rbf_svm(
D, L, C=5, gamma=0.1, model_name='final_surf_30_svm')
print('Time spend: {:.2f} s'.format(time.time() - temp))
temp = time.time()
# Read the test set
test_images_filenames, test_labels = io.load_test_set()
print('Loaded {} test images.'.format(len(test_images_filenames)))
# Feature extraction with surf, prediction with SVM and aggregation to obtain final class
print('Predicting test data...')
result = joblib.Parallel(n_jobs=N_JOBS, backend='threading')(
joblib.delayed(parallel_testing)(test_image, test_label, lin_svm,
std_scaler, None)
for test_image, test_label in zip(test_images_filenames, test_labels))
print('Sweeping ' + sweep_mode + ' for kernel type ' + kernel + '...')
for p1 in sw1:
for p2 in sw2:
print('p1 value ' + str(p1) + ' and p2 value ' + str(p2))
std_scaler = None
pca = None
if sweep_mode == 'cost':
if kernel == 'linear':
svm, std_scaler, pca = classification.train_linear_svm(
D, L, p1, dim_reduction=23)
elif kernel == 'poly':
svm, std_scaler, pca = classification.train_poly_svm(
D, L, p1, dim_reduction=23)
elif kernel == 'rbf':
svm, std_scaler, pca = classification.train_rbf_svm(
D, L, p1, dim_reduction=23)
elif kernel == 'sigmoid':
svm, std_scaler, pca = classification.train_sigmoid_svm(
D, L, p1, dim_reduction=23)
else:
svm, std_scaler, pca = classification.train_linear_svm(
D, L, p1, dim_reduction=23)
print('Time spend: {:.2f} s'.format(time.time() - temp))
temp = time.time()
elif sweep_mode == 'params':
if kernel == 'poly':
svm, std_scaler, pca = classification.train_poly_svm(
D, L, degree=p1, coef0=p2)
elif kernel == 'rbf':
svm, std_scaler, pca = classification.train_rbf_svm(
D, L, gamma=p1)
print('Elapsed time: {:.2f} s'.format(time.time() - start))
temp = time.time()
print('Creating codebook with {} visual words'.format(K))
codebook = bovw.create_codebook(D, codebook_name='dense_codebook')
print('Elapsed time: {:.2f} s'.format(time.time() - temp))
temp = time.time()
print('Getting visual words from training set...')
vis_words, labels = bovw.visual_words(D, L, I, codebook)
print('Elapsed time: {:.2f} s'.format(time.time() - temp))
temp = time.time()
# Train Linear SVM classifier
print('Training the SVM classifier...')
lin_svm, std_scaler, pca = classification.train_rbf_svm(vis_words, labels, C=7.76, gamma=0.0013808, dim_reduction=None)
print('Elapsed time: {:.2f} s'.format(time.time() - temp))
temp = time.time()
# Read the test set
test_images_filenames, test_labels = io.load_test_set()
print('Loaded {} test images.'.format(len(test_images_filenames)))
# Feature extraction with sift, prediction with SVM and aggregation to obtain final class
print('Predicting test data...')
test_results = joblib.Parallel(n_jobs=N_JOBS, backend='threading')(
joblib.delayed(parallel_testing)(test_image, test_label, codebook, lin_svm, std_scaler, pca) for
test_image, test_label in
zip(test_images_filenames, test_labels))
pred_results = [x[0] for x in test_results]
print('Elapsed time: {:.2f} s'.format(time.time() - start))
temp = time.time()
print('Creating codebook with {} visual words'.format(K))
codebook = bovw.create_codebook(D, codebook_name='codebook_256_dense')
print('Elapsed time: {:.2f} s'.format(time.time() - temp))
temp = time.time()
print('Getting visual words from training set...')
vis_words, labels = bovw.visual_words(D, L, I, codebook)
print('Elapsed time: {:.2f} s'.format(time.time() - temp))
temp = time.time()
# Train Linear SVM classifier
print('Training the SVM classifier...')
lin_svm, std_scaler, pca = classification.train_rbf_svm(vis_words, labels, C=1.65, gamma=0.005926, dim_reduction=None)
print('Elapsed time: {:.2f} s'.format(time.time() - temp))
temp = time.time()
# Read the test set
test_images_filenames, test_labels = io.load_test_set()
print('Loaded {} test images.'.format(len(test_images_filenames)))
# Feature extraction with sift, prediction with SVM and aggregation to obtain final class
print('Predicting test data...')
test_results = joblib.Parallel(n_jobs=N_JOBS, backend='threading')(
joblib.delayed(parallel_testing)(test_image, test_label, codebook, lin_svm, std_scaler, pca) for
test_image, test_label in
zip(test_images_filenames, test_labels))
pred_results = [x[0] for x in test_results]
# Read the training set
train_images_filenames, train_labels = io.load_training_set()
print('Loaded {} train images.'.format(len(train_images_filenames)))
# Feature extraction with sift
print('Obtaining sift features...')
D, L, _, _ = feature_extraction.parallel_sift(train_images_filenames,
train_labels,
n_jobs=N_JOBS)
print('Time spend: {:.2f} s'.format(time.time() - start))
temp = time.time()
# Train Linear SVM classifier
print('Training the SVM classifier...')
lin_svm, std_scaler, pca = classification.train_rbf_svm(
D, L, dim_reduction=23, model_name='final_sift_all_svmPCA')
print('Time spend: {:.2f} s'.format(time.time() - temp))
temp = time.time()
# Read the test set
test_images_filenames, test_labels = io.load_test_set()
print('Loaded {} test images.'.format(len(test_images_filenames)))
# Feature extraction with sift, prediction with SVM and aggregation to obtain final class
print('Predicting test data...')
result = joblib.Parallel(n_jobs=N_JOBS, backend='threading')(
joblib.delayed(parallel_testing)(test_image, test_label, lin_svm,
std_scaler, pca)
for test_image, test_label in zip(test_images_filenames, test_labels))
# Feature extraction
print('Obtaining features...')
D, L, _ = train_function(train_images_filenames,
train_labels,
num_samples_class=num_samples,
n_jobs=N_JOBS)
print('Time spend: {:.2f} s'.format(time.time() - start))
temp = time.time()
# Train Linear SVM classifier
print('Training the SVM with RBF kernel classifier...')
svm, std_scaler, pca = classification.train_rbf_svm(
D,
L,
C=5,
gamma=0.1,
dim_reduction=dim_red_option,
model_name='svm_{}_{}s_{}c'.format(
fe_name, num_samples if num_samples > -1 else 'all',
dim_red_option
if dim_red_option is not None else 'all'))
print('Time spend: {:.2f} s'.format(time.time() - temp))
temp = time.time()
# Feature extraction with sift, prediction with SVM and aggregation to obtain final class
print('Predicting test data...')
result = joblib.Parallel(n_jobs=N_JOBS, backend='threading')(
joblib.delayed(predict_function)(test_image, test_label,
svm, std_scaler, pca)
for test_image, test_label in zip(test_images_filenames,
test_labels))
print('Time spend: {:.2f} s'.format(time.time() - temp))
# Feature extraction with sift
print('Obtaining sift features...')
D, L, _, _ = feature_extraction.parallel_sift(train_images_filenames,
train_labels,
num_samples_class=30,
n_jobs=N_JOBS)
print('Time spend: {:.2f} s'.format(time.time() - start))
temp = time.time()
# Train Linear SVM classifier
print('Training the SVM classifier...')
lin_svm, std_scaler, pca = classification.train_rbf_svm(
D,
L,
C=5,
gamma=0.1,
model_name='final_sift_30_svm_pca23',
dim_reduction=23)
print('Time spend: {:.2f} s'.format(time.time() - temp))
temp = time.time()
# Read the test set
test_images_filenames, test_labels = io.load_test_set()
print('Loaded {} test images.'.format(len(test_images_filenames)))
# Feature extraction with sift, prediction with SVM and aggregation to obtain final class
print('Predicting test data...')
result = joblib.Parallel(n_jobs=N_JOBS, backend='threading')(
joblib.delayed(parallel_testing)(test_image, test_label, lin_svm,
