#time.sleep(1)
curTime = time.time() # calc fps
find_results = []
frame = frame[:, :, 0:3]
boxes, scores = face_detector.detect(frame)
face_boxes = boxes[np.argwhere(scores>0.3).reshape(-1)]
face_scores = scores[np.argwhere(scores>0.3).reshape(-1)]
print('Detected_FaceNum: %d' % len(face_boxes))
if len(face_boxes) > 0:
for i in range(len(face_boxes)):
box = face_boxes[i]
cropped_face = frame[box[0]:box[2], box[1]:box[3], :]
cropped_face = cv2.resize(cropped_face, (160, 160), interpolation=cv2.INTER_AREA)
feature = face_recognition.recognize(cropped_face)
name = face_classfier.classify(feature)
cv2.rectangle(frame, (box[1], box[0]), (box[3], box[2]), (0, 255, 0), 2)
# plot result idx under box
text_x = box[1]
text_y = box[2] + 20
cv2.putText(frame, name, (text_x, text_y), cv2.FONT_HERSHEY_COMPLEX_SMALL,
1, (0, 0, 255), thickness=1, lineType=2)
else:
print('Unable to align')
sec = curTime - prevTime
prevTime = curTime
fps = 1 / (sec)
curTime = time.time() # calc fps
find_results = []
frame = frame[:, :, 0:3]
if skip==15:
boxes, scores = face_detector.detect(frame)
face_boxes = boxes[np.argwhere(scores>0.3).reshape(-1)]
face_scores = scores[np.argwhere(scores>0.3).reshape(-1)]
print('Detected_FaceNum: %d' % len(face_boxes))
if len(face_boxes) > 0:
for i in range(len(face_boxes)):
box = face_boxes[i]
cropped_face = frame[box[0]:box[2], box[1]:box[3], :]
cropped_face = cv2.resize(cropped_face, (160, 160), interpolation=cv2.INTER_AREA)
feature = face_recognition.recognize(cropped_face,mobilenet=mobilenet)
print("len features {}".format(len(feature)))
ind,probab = face_classfier.ensemble(feature)
name=index_to_name(ind)
print(name,probab)
'''sum=np.load('sum.npy')
print(np.shape(sum))
names=['ayush','gautham','lokesh','milind','nishant','pranjal','rishhanth','sumanth']
diff=sum-feature
res= np.linalg.norm(diff,axis=1)
print(res)
res_ind=np.argmin(res)
if res[res_ind] <= 0.5 :
y_pred = names[res_ind]'''
import cv2
import numpy as np
from recognition.FaceRecognition import FaceRecognition
from detection.FaceDetector import FaceDetector
face_detector = FaceDetector()
face_recognition = FaceRecognition()
image_files = ['./media/1.jpg', './media/2.jpg']
for input_str in image_files:
img = cv2.imread(input_str)
boxes, scores = face_detector.detect(img)
face_boxes = boxes[np.argwhere(scores > 0.5).squeeze()]
print('Number of face in image:', len(face_boxes))
for box in face_boxes:
cropped_face = img[box[0]:box[2], box[1]:box[3], :]
cropped_face = cv2.resize(cropped_face, (160, 160),
interpolation=cv2.INTER_AREA)
print('Face descriptor:')
print(face_recognition.recognize(cropped_face), '\n')
cv2.imshow('image', cropped_face)
cv2.waitKey(0)
cv2.destroyAllWindows()
def main(args):
face_detector = FaceDetector()
face_recognition = FaceRecognition(args.model)
face_classfier = FaceClassifier(args.classifier_filename)
video_capture = cv2.VideoCapture(args.video_input)
output_file = './media/result/' + os.path.basename(
args.video_input) + '_result.avi'
fourcc = cv2.VideoWriter_fourcc(*'XVID')
out = cv2.VideoWriter(
output_file, fourcc, 24.0,
(int(video_capture.get(3)), int(video_capture.get(4))))
print('Start Recognition!')
prevTime = 0
while video_capture.isOpened():
ret, frame = video_capture.read()
curTime = time.time() # calc fps
find_results = []
frame = frame[:, :, 0:3]
boxes, scores = face_detector.detect(frame)
face_boxes = boxes[np.argwhere(scores > 0.3).reshape(-1)]
face_scores = scores[np.argwhere(scores > 0.3).reshape(-1)]
print('Detected_FaceNum: %d' % len(face_boxes))
if len(face_boxes) > 0:
for i in range(len(face_boxes)):
box = face_boxes[i]
cropped_face = frame[box[0]:box[2], box[1]:box[3], :]
cropped_face = cv2.resize(cropped_face, (160, 160),
interpolation=cv2.INTER_AREA)
feature = face_recognition.recognize(cropped_face)
name = face_classfier.classify(feature)
cv2.rectangle(frame, (box[1], box[0]), (box[3], box[2]),
(0, 255, 0), 2)
# plot result idx under box
text_x = box[1]
text_y = box[2] + 20
cv2.putText(frame,
name, (text_x, text_y),
cv2.FONT_HERSHEY_COMPLEX_SMALL,
1, (0, 0, 255),
thickness=1,
lineType=2)
else:
print('Unable to align')
sec = curTime - prevTime
prevTime = curTime
fps = 1 / (sec)
str = 'FPS: %2.3f' % fps
text_fps_x = len(frame[0]) - 150
text_fps_y = 20
cv2.putText(frame,
str, (text_fps_x, text_fps_y),
cv2.FONT_HERSHEY_COMPLEX_SMALL,
1, (0, 0, 0),
thickness=1,
lineType=2)
out.write(frame)
video_capture.release()
out.release()
cv2.destroyAllWindows()
print('Number of classes: %d' % len(dataset))
print('Number of images: %d' % len(paths))
# Run forward pass to calculate embeddings
print('Calculating features for images')
image_size = 160
nrof_images = len(paths)
features = np.zeros((2 * nrof_images, 512))
labels = np.asarray(labels).repeat(2)
for i in range(nrof_images):
img = cv2.imread(paths[i])
if img is None:
print('Open image file failed: ' + paths[i])
continue
boxes, scores = face_detector.detect(img)
if len(boxes) < 0 or scores[0] < 0.5:
print('No face found in ' + paths[i])
continue
cropped_face = img[boxes[0][0]:boxes[0][2], boxes[0][1]:boxes[0][3], :]
cropped_face_flip = cv2.flip(cropped_face, 1)
features[2 * i, :] = face_recognition.recognize(cropped_face)
features[2 * i + 1, :] = face_recognition.recognize(cropped_face_flip)
print('Start training for images')
face_classfier.train(features,
labels,
model='svm',
save_model_path='./classifier/trained_classifier.pkl')
nrof_images = len(paths)
features = np.zeros((2 * nrof_images, args.embedding_size))
labels = np.asarray(labels).repeat(2)
for i in range(nrof_images):
img = cv2.imread(paths[i])
if img is None:
print('Open image file failed: ' + paths[i])
continue
boxes, scores = face_detector.detect(img)
if len(boxes) < 0 or scores[0] < 0.5:
print('No face found in ' + paths[i])
continue
cropped_face = img[boxes[0][0]:boxes[0][2], boxes[0][1]:boxes[0][3], :]
cropped_face_flip = cv2.flip(cropped_face, 1)
features[2 * i, :] = face_recognition.recognize(cropped_face,
mobilenet=mobilenet)
features[2 * i + 1, :] = face_recognition.recognize(cropped_face_flip,
mobilenet=mobilenet)
np.save('features', features)
np.save('labels', labels)
'''trained on 8 classes . 1110 images . accuracy:0.9617117117117117
precision,accuracy and f1 score is (array([0.98387097, 1. , 0.82461538, 0.9887218 , 0.98245614,
0.9858156 , 0.95955882, 0.99618321]), array([0.87142857, 1. , 0.97101449, 0.99621212, 0.9929078 ,
0.99285714, 0.93884892, 0.93214286]), array([0.92424242, 1. , 0.89184692, 0.99245283, 0.98765432,
0.98932384, 0.94909091, 0.96309963]), array([280, 280, 276, 264, 282, 280, 278, 280]))'''
print('Start training for images')
face_classfier.train(features,
labels,
model=model_type,