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='default_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, spatial_pyramid=True)
print('Elapsed time: {:.2f} s'.format(time.time() - temp))
temp = time.time()
# Train Linear SVM classifier
print('Training the SVM classifier...')
pyramid_svm, std_scaler, pca = classification.train_pyramid_svm(vis_words, labels, C=0.0004, dim_reduction=None, model_name='svm_pyramid_dense' )
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, pyramid_svm, std_scaler, pca) for
test_image, test_label in
print('Creating codebook with {} visual words'.format(settings.codebook_size))
#gmm = bovw.create_gmm(D, codebook_name='gmm_{}_dense'.format(settings.codebook_size))
codebook = bovw.create_codebook(D)
print('Elapsed time: {:.2f} s'.format(time.time() - temp))
temp = time.time()
print('Getting visual words from training set...')
#fisher, labels = bovw.fisher_vectors(D, L, I, gmm)
pyramid, labels = bovw.visual_words(D, L, I, codebook, spatial_pyramid=True)
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, _ = classification.train_pyramid_svm(pyramid, train_labels, C=0.0009, 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=settings.n_jobs, backend='threading')(
joblib.delayed(parallel_testing)(test_image, test_label, lin_svm, std_scaler) for
test_image, test_label in
zip(test_images_filenames, test_labels))
pred_results = [x[0] for x in test_results]
codebook = bovw.create_codebook(D) #, codebook_name='default_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,
spatial_pyramid=True)
print('Elapsed time: {:.2f} s'.format(time.time() - temp))
temp = time.time()
# Train Linear SVM classifier
print('Training the SVM classifier...')
pyramid_svm, std_scaler, pca = classification.train_pyramid_svm(
vis_words, labels, C=0.0009, 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=settings.n_jobs,
backend='threading')(joblib.delayed(parallel_testing)(
test_image, test_label, codebook, pyramid_svm, std_scaler, pca)
for test_image, test_label in zip(
print('Getting visual words from training set...')
vis_words, labels = bovw.visual_words(D,
L,
I,
codebook,
spatial_pyramid=True,
normalization='l1')
print('Elapsed time: {:.2f} s'.format(time.time() - temp))
temp = time.time()
# Train Linear SVM classifier
print('Training the SVM classifier...')
pyramid_svm, std_scaler, pca = classification.train_pyramid_svm(
vis_words,
labels,
C=840.42,
dim_reduction=None,
model_name='svm_pyramid_l1')
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,