def predict_top5(self, video_file_path):
x = extract_vgg16_features_live(self.vgg16_model, video_file_path)
frames = x.shape[0]
if frames > self.expected_frames:
x = x[0:self.expected_frames, :]
elif frames < self.expected_frames:
temp = np.zeros(shape=(self.expected_frames, x.shape[1]))
temp[0:frames, :] = x
x = temp
predict = self.model.predict(np.array([x]))
print(predict)
predicted_class = np.argsort(predict[0])[-5:]
print(predicted_class)
predict_labels = []
#top_values= [class_prob[i] for i in np.argsort(class_prob)[-5:]]
for _class in predicted_class:
#print(_class)
label = self.labels_idx2word[_class]
predict_labels.append(label)
#predicted_label = labels_idx2word_top5(predicted_class)
#predicted_class = np.argmax(self.model.predict(np.array([x]))[:5])
#predicted_label = self.labels_idx2word[predicted_class]
print(predict_labels)
return predict_labels
def predict(self, video_file_path):
x = extract_vgg16_features_live(self.vgg16_model, video_file_path)
frames = x.shape[0]
if frames > self.expected_frames:
x = x[0:self.expected_frames, :]
elif frames < self.expected_frames:
temp = np.zeros(shape=(self.expected_frames, x.shape[1]))
temp[0:frames, :] = x
x = temp
predicted_class = np.argmax(self.model.predict(np.array([x]))[0])
predicted_label = self.labels_idx2word[predicted_class]
return predicted_label
def predict(self, video_file_path , from_picture=False):
if from_picture:
x = extract_vgg16_features_live_pictures(self.vgg16_model, video_file_path)
else:
x = extract_vgg16_features_live(self.vgg16_model, video_file_path)
frames = x.shape[0]
if frames > self.expected_frames:
x = x[0:self.expected_frames, :]
elif frames < self.expected_frames:
temp = np.zeros(shape=(self.expected_frames, x.shape[1]))
temp[0:frames, :] = x
x = temp
yourlist = self.model.predict(np.array([x]))[0]
print("%.5f " * len(yourlist) % tuple(yourlist))
predicted_class = np.argmax(self.model.predict(np.array([x]))[0])
predicted_label = self.labels_idx2word[predicted_class]
return predicted_label
def predict(self, video_file_path, interval=0, vgg16_include_top=True):
predicted_labels = []
feature_dir_name = video_file_path + '-VGG16-Features'
if not vgg16_include_top:
feature_dir_name = video_file_path + '-VGG16-HiDimFeatures'
if not os.path.exists(feature_dir_name):
os.makedirs(feature_dir_name)
if (interval != 0):
print("Predicting each interval")
features_arr = extract_vgg16_features_live_each_interval(
self.vgg16_model, video_file_path, feature_dir_name, interval)
for x in features_arr:
frames = x.shape[0]
if frames > self.expected_frames:
x = x[0:self.expected_frames, :]
elif frames < self.expected_frames:
temp = np.zeros(shape=(self.expected_frames, x.shape[1]))
temp[0:frames, :] = x
x = temp
predicted_class = np.argmax(
self.model.predict(np.array([x]))[0])
predicted_label = self.labels_idx2word[predicted_class]
predicted_labels.append(predicted_label)
return predicted_labels
else:
print("Predicting entire video")
x = extract_vgg16_features_live(self.vgg16_model, video_file_path,
feature_dir_name)
frames = x.shape[0]
if frames > self.expected_frames:
x = x[0:self.expected_frames, :]
elif frames < self.expected_frames:
temp = np.zeros(shape=(self.expected_frames, x.shape[1]))
temp[0:frames, :] = x
x = temp
predicted_class = np.argmax(self.model.predict(np.array([x]))[0])
predicted_label = self.labels_idx2word[predicted_class]
return predicted_label
def predict(self, video_file_path, label, correct_count, count):
return extract_vgg16_features_live(self.vgg16_model, video_file_path, self.model, self.expected_frames, self.labels_idx2word, label, correct_count, count)
"""
