def update_buffer(self):
detection_model_path = './Engine/trained_models/detection_models/haarcascade_frontalface_default.xml'
face_detection = load_detection_model(detection_model_path)
# hyper-parameters for bounding boxes shape
frame_window = 10
emotion_offsets = (20, 40)
emotion_labels = get_labels('fer2013')
# starting lists for calculating modes
emotion_window = []
while True:
for element in self.VideoStreamer_list:
if element[1].more():
bgr_image = element[1].read()
# sleep(1/30)
# bgr_image = trim_frame(bgr_image)
gray_image = cv2.cvtColor(bgr_image, cv2.COLOR_BGR2GRAY)
faces = detect_faces(face_detection, gray_image)
# print(str(faces))
for face_coordinates in faces:
x1, x2, y1, y2 = apply_offsets(face_coordinates,
emotion_offsets)
gray_face = gray_image[y1:y2, x1:x2]
try:
gray_face = cv2.resize(gray_face,
(self.emotion_target_size))
except:
continue
gray_face = preprocess_input(gray_face, True)
gray_face = np.expand_dims(gray_face, 0)
gray_face = np.expand_dims(gray_face, -1)
with self.graph.as_default():
emotion_prediction = self.emotion_classifier.predict(
gray_face)
percentage = np.amax(emotion_prediction)
emotion_label_arg = np.argmax(emotion_prediction)
emotion_text = emotion_labels[emotion_label_arg]
if not self.Q.full():
self.Q.put([element[0], emotion_text, percentage])
else:
continue
else:
continue
from Engine.utils.inference import draw_text
from Engine.utils.inference import draw_bounding_box
from Engine.utils.inference import apply_offsets
from Engine.utils.inference import load_detection_model
from Engine.utils.preprocessor import preprocess_input
detection_model_path = './Engine/trained_models/detection_models/haarcascade_frontalface_default.xml'
emotion_model_path = './Engine/trained_models/emotion_models/fer2013_mini_XCEPTION.102-0.66.hdf5'
emotion_labels = get_labels('fer2013')
frame_window = 5
emotion_offsets = (20, 40)
# loading models
face_detection = load_detection_model(detection_model_path)
emotion_classifier = load_model(emotion_model_path, compile=False)
# getting input model shapes for inference
emotion_target_size = emotion_classifier.input_shape[1:3]
# starting lists for calculating modes
emotion_window = []
# starting video streaming
cv2.namedWindow('window_frame')
vs = Videostream('https://www.twitch.tv/mrmcpowned', resolution='720p60')
while True:
def update_buffer(self):
detection_model_path = './Engine/trained_models/detection_models/haarcascade_frontalface_default.xml'
face_detection = load_detection_model(detection_model_path)
# hyper-parameters for bounding boxes shape
frame_window = 10
emotion_offsets = (20, 40)
emotion_labels = get_labels('fer2013')
# starting lists for calculating modes
emotion_window = []
while True:
if self.VideoStreamer.more():
bgr_image = self.VideoStreamer.read()
# sleep(1/30)
# bgr_image = trim_frame(bgr_image)
gray_image = cv2.cvtColor(bgr_image, cv2.COLOR_BGR2GRAY)
rgb_image = cv2.cvtColor(bgr_image, cv2.COLOR_BGR2RGB)
faces = detect_faces(face_detection, gray_image)
# print(str(faces))
for face_coordinates in faces:
x1, x2, y1, y2 = apply_offsets(face_coordinates,
emotion_offsets)
gray_face = gray_image[y1:y2, x1:x2]
try:
gray_face = cv2.resize(gray_face,
(self.emotion_target_size))
except:
continue
gray_face = preprocess_input(gray_face, True)
gray_face = np.expand_dims(gray_face, 0)
gray_face = np.expand_dims(gray_face, -1)
with self.graph.as_default():
emotion_prediction = self.emotion_classifier.predict(
gray_face)
# print(str(emotion_prediction))
# angry = emotion_prediction[0][0]
# disgust = emotion_prediction[0][1]
# fear = emotion_prediction[0][2]
# happy = emotion_prediction[0][3]
# sad = emotion_prediction[0][4]
# surprise = emotion_prediction[0][5]
# neutral = emotion_prediction[0][6]
# with open('../emotion.txt', 'a') as f:
# f.write('{},{},{},{},{},{},{}\n'.format(angry, disgust, fear, happy, sad, surprise, neutral))
emotion_probability = np.max(emotion_prediction)
emotion_label_arg = np.argmax(emotion_prediction)
emotion_text = emotion_labels[emotion_label_arg]
emotion_window.append(emotion_text)
if len(emotion_window) > frame_window:
emotion_window.pop(1)
try:
emotion_mode = mode(emotion_window)
except:
continue
if emotion_text == 'angry':
color = emotion_probability * np.asarray((255, 0, 0))
elif emotion_text == 'sad':
color = emotion_probability * np.asarray((0, 0, 255))
elif emotion_text == 'happy':
color = emotion_probability * np.asarray((255, 255, 0))
elif emotion_text == 'surprise':
color = emotion_probability * np.asarray((0, 255, 255))
else:
color = emotion_probability * np.asarray((0, 255, 0))
color = color.astype(int)
color = color.tolist()
draw_bounding_box(face_coordinates, rgb_image, color)
draw_text(face_coordinates, rgb_image, emotion_mode, color,
0, -45, 1, 1)
bgr_image = cv2.cvtColor(rgb_image, cv2.COLOR_RGB2BGR)
if not self.Q.full():
self.Q.put(bgr_image)
else:
continue
else:
continue