def get_chosen_costs(opts, num_classes):
costs_list = svm_helper.parse_cost_list(opts.costs_list)
train_ap_matrix = np.zeros((num_classes, len(costs_list)))
for cls in range(num_classes):
for cost_idx in range(len(costs_list)):
cost = costs_list[cost_idx]
_, ap_out_file = svm_helper.get_svm_train_output_files(
cls, cost, opts.output_path
)
train_ap_matrix[cls][cost_idx] = float(
np.load(ap_out_file, encoding='latin1')[0]
)
argmax_cls = np.argmax(train_ap_matrix, axis=1)
chosen_cost = [costs_list[idx] for idx in argmax_cls]
logger.info('chosen_cost: {}'.format(chosen_cost))
np.save(
os.path.join(opts.output_path, 'crossval_ap.npy'),
np.array(train_ap_matrix)
)
np.save(
os.path.join(opts.output_path, 'chosen_cost.npy'),
np.array(chosen_cost)
)
logger.info('saved crossval_ap AP to file: {}'.format(
os.path.join(opts.output_path, 'crossval_ap.npy')))
logger.info('saved chosen costs to file: {}'.format(
os.path.join(opts.output_path, 'chosen_cost.npy')))
return np.array(chosen_cost)
def task(cls, cost, opts, features, targets):
out_file, ap_out_file = svm_helper.get_svm_train_output_files(
cls, cost, opts.output_path)
if not (os.path.exists(out_file) and os.path.exists(ap_out_file)):
clf = LinearSVC(
C=cost,
class_weight={
1: 2,
-1: 1
},
intercept_scaling=1.0,
verbose=0,
penalty='l2',
loss='squared_hinge',
tol=0.0001,
dual=True,max_iter = 500,
#max_iter=2000,
)
cls_labels = targets[:, cls].astype(dtype=np.int32, copy=True)
cls_labels[np.where(cls_labels == 0)] = -1
ap_scores = cross_val_score(
clf, features, cls_labels, cv=3, scoring='average_precision')
clf.fit(features, cls_labels)
np.save(ap_out_file, np.array([ap_scores.mean()]))
with open(out_file, 'wb') as fwrite:
pickle.dump(clf, fwrite)
return 0
def task(cls, cost, opts, features, targets):
out_file, ap_out_file = svm_helper.get_svm_train_output_files(
cls, cost, opts.output_path)
if os.path.exists(out_file) and os.path.exists(ap_out_file):
logger.info('SVM model exists: {}'.format(out_file))
logger.info('AP file exists: {}'.format(ap_out_file))
else:
#logger.info('Training model with the cost: {}'.format(cost))
clf = LinearSVC(
C=cost,
class_weight={
1: 2,
-1: 1
},
intercept_scaling=1.0,
verbose=1,
penalty='l2',
loss='squared_hinge',
tol=0.0001,
dual=True,
max_iter=2000,
)
cls_labels = targets[:, cls].astype(dtype=np.int32, copy=True)
# meaning of labels in VOC/COCO original loaded target files:
# label 0 = not present, set it to -1 as svm train target
# label 1 = present. Make the svm train target labels as -1, 1.
cls_labels[np.where(cls_labels == 0)] = -1
num_positives = len(np.where(cls_labels == 1)[0])
num_negatives = len(cls_labels) - num_positives
#logger.info('cls: {} has +ve: {} -ve: {} ratio: {}'.format(
# cls, num_positives, num_negatives,
# float(num_positives) / num_negatives)
#)
#logger.info('features: {} cls_labels: {}'.format(
# features.shape, cls_labels.shape))
ap_scores = cross_val_score(clf,
features,
cls_labels,
cv=3,
scoring='average_precision')
clf.fit(features, cls_labels)
#logger.info('cls: {} cost: {} AP: {} mean:{}'.format(
# cls, cost, ap_scores, ap_scores.mean()))
#logger.info('Saving cls cost AP to: {}'.format(ap_out_file))
np.save(ap_out_file, np.array([ap_scores.mean()]))
#logger.info('Saving SVM model to: {}'.format(out_file))
with open(out_file, 'wb') as fwrite:
pickle.dump(clf, fwrite)
return 0
def train_svm(opts):
assert os.path.exists(opts.data_file), "Data file not found. Abort!"
if not os.path.exists(opts.output_path):
os.makedirs(opts.output_path)
features, targets = svm_helper.load_input_data(
opts.data_file, opts.targets_data_file
)
# normalize the features: N x 9216 (example shape)
features = svm_helper.normalize_features(features)
# parse the cost values for training the SVM on
costs_list = svm_helper.parse_cost_list(opts.costs_list)
logger.info('Training SVM for costs: {}'.format(costs_list))
# classes for which SVM training should be done
if opts.cls_list:
cls_list = [int(cls) for cls in opts.cls_list.split(",")]
else:
num_classes = targets.shape[1]
cls_list = range(num_classes)
logger.info('Training SVM for classes: {}'.format(cls_list))
for cls_idx in range(len(cls_list)):
cls = cls_list[cls_idx]
for cost_idx in range(len(costs_list)):
cost = costs_list[cost_idx]
out_file, ap_out_file = svm_helper.get_svm_train_output_files(
cls, cost, opts.output_path
)
if os.path.exists(out_file) and os.path.exists(ap_out_file):
logger.info('SVM model exists: {}'.format(out_file))
logger.info('AP file exists: {}'.format(ap_out_file))
else:
logger.info('Training model with the cost: {}'.format(cost))
clf = LinearSVC(
C=cost, class_weight={1: 2, -1: 1}, intercept_scaling=1.0,
verbose=1, penalty='l2', loss='squared_hinge', tol=0.0001,
dual=True, max_iter=2000,
)
cls_labels = targets[:, cls].astype(dtype=np.int32, copy=True)
# meaning of labels in VOC/COCO original loaded target files:
# label 0 = not present, set it to -1 as svm train target
# label 1 = present. Make the svm train target labels as -1, 1.
cls_labels[np.where(cls_labels == 0)] = -1
num_positives = len(np.where(cls_labels == 1)[0])
num_negatives = len(cls_labels) - num_positives
logger.info('cls: {} has +ve: {} -ve: {} ratio: {}'.format(
cls, num_positives, num_negatives,
float(num_positives) / num_negatives)
)
logger.info('features: {} cls_labels: {}'.format(
features.shape, cls_labels.shape))
ap_scores = cross_val_score(
clf, features, cls_labels, cv=3, scoring='average_precision'
)
clf.fit(features, cls_labels)
logger.info('cls: {} cost: {} AP: {} mean:{}'.format(
cls, cost, ap_scores, ap_scores.mean()))
logger.info('Saving cls cost AP to: {}'.format(ap_out_file))
np.save(ap_out_file, np.array([ap_scores.mean()]))
logger.info('Saving SVM model to: {}'.format(out_file))
with open(out_file, 'wb') as fwrite:
pickle.dump(clf, fwrite)