def by_dataset(self, vid_path='./data/freeman1/'):
assert os.path.exists(vid_path)
gt_boxes = BoundingBox.read_vid_gt(vid_path)
curr_bbox = None
self.stopwatch.start('total')
_logger.info('---- start dataset l=%d' % (len(gt_boxes)))
for idx, gt_box in enumerate(gt_boxes):
img = commons.imread(
os.path.join(vid_path, 'img', '%04d.jpg' % (idx + 1)))
self.imgwh = Coordinate.get_imgwh(img)
if idx == 0:
# initialization : initial fine-tuning
self.initial_finetune(img, gt_box)
curr_bbox = gt_box
# tracking
predicted_box = self.tracking(img, curr_bbox)
self.show(img, gt_box=gt_box, predicted_box=predicted_box)
# cv2.imwrite('/Users/ildoonet/Downloads/aaa/%d.jpg' % self.iteration, img)
curr_bbox = predicted_box
self.stopwatch.stop('total')
_logger.info('----')
_logger.info(self.stopwatch)
_logger.info('%.3f FPS' %
(len(gt_boxes) / self.stopwatch.get_elapsed('total')))
def __call__(self, results):
filename = osp.join(results['img_prefix'],
results['img_info']['filename'])
img = commons.imread(filename)
if self.to_float32:
img = img.astype(np.float32)
results['filename'] = filename
results['img'] = img
results['img_shape'] = img.shape
results['ori_shape'] = img.shape
return results
def show(self, img, delay=1, predicted_box=None, gt_box=None):
if isinstance(img, str):
img = commons.imread(img)
if gt_box is not None:
gt_box.draw(img, BoundingBox.COLOR_GT)
if predicted_box is not None:
predicted_box.draw(img, BoundingBox.COLOR_PREDICT)
visualizer.image('result', img)
# cv2.imshow('result', img)
cv2.waitKey(delay)
def show(self, img,idx, delay=1, predicted_box=None, gt_box=None):
if isinstance(img, str):
img = commons.imread(img)
if gt_box is not None:
gt_box.draw(img, BoundingBox.COLOR_GT)
if predicted_box is not None:
predicted_box.draw(img, BoundingBox.COLOR_PREDICT)
cv2.imshow('result', img)
path = './result/'
os.path.join(path)
cv2.imwrite( '%06d.jpg' % (idx + 1), img)
cv2.waitKey(delay)
def __call__(self, results):
filename1 = osp.join(results['img_prefix'],
results['img_info']['filename1'])
img1 = commons.imread(filename1)
filename2 = osp.join(results['img_prefix'],
results['img_info']['filename2'])
img2 = commons.imread(filename2)
if self.to_float32:
img1 = img1.astype(np.float32)
img2 = img2.astype(np.float32)
results['filename1'] = filename1
results['img1'] = img1
results['img1_shape'] = img1.shape
results['ori1_shape'] = img1.shape
results['filename2'] = filename2
results['img2'] = img2
results['img2_shape'] = img2.shape
results['ori2_shape'] = img2.shape
return results
def img_label(lb_dir_dict):
img_label = {}
train_img = {}
test_img = {}
for key in lb_dir_dict:
imgs = []
img_dir = glob.glob(lb_dir_dict[key]+'/*.bmp')
for idx, dir in enumerate(img_dir):
print("Read: ", dir)
img = commons.imread(dir)
img = commons.extract_region(img)
imgs.append(img)
img_label[key] = imgs
train_size = int(len(imgs)*0.8)
train_img[key] = imgs[:train_size]
test_img[key] = imgs[train_size:]
return img_label, train_img, test_img
def by_dataset(self, vid_path='./data/LT52/'):
"""
'./data/BlurCar2/'
LT52
"""
assert os.path.exists(vid_path)
id=0
gt_boxes = BoundingBox.read_vid_gt(vid_path,id)
print("gt_boxes>>", gt_boxes)
curr_bbox = None
self.stopwatch.start('total')
_logger.info('---- start dataset l=%d' % (len(gt_boxes)))
for idx, gt_box in enumerate(gt_boxes):
print('\nimage number : %04d.jpg' %(idx+1))
print('vid path is %s' %vid_path)
if idx!=1:
img = commons.imread(os.path.join(vid_path, 'img', '%04d.jpeg' % (idx + 1)))
self.imgwh = Coordinate.get_imgwh(img)
if idx == 0:
# initialization : initial fine-tuning
# print("gt_box >", gt_box)
self.initial_finetune(img, gt_box)
curr_bbox = gt_box
# tracking
# print("curr_bbox>>", curr_bbox, idx)
predicted_box = self.tracking(img, curr_bbox,idx)
# print("predicted_box>>", predicted_box, predicted_box.xy.x, predicted_box.xy.y)
self.show(img, idx,gt_box=gt_box, predicted_box=predicted_box)
# cv2.imwrite('/Users/ildoonet/Downloads/aaa/%d.jpg' % self.iteration, img)
curr_bbox = predicted_box
self.stopwatch.stop('total')
_logger.info('----')
_logger.info(self.stopwatch)
_logger.info('%.3f FPS' % (len(gt_boxes) / self.stopwatch.get_elapsed('total')))
from conf.configs import ADNetConf
ADNetConf.get('./conf/dylan.yaml')
input_node = tf.placeholder(tf.float32, shape=(None, 2, 107, 107, 3), name='patch')
tensor_lb_action = tf.placeholder(tf.float32, shape=(None, 4), name='lb_action') # actions
tensor_lb_class = tf.placeholder(tf.int32, shape=(None, ), name='lb_class') # 2 actions
is_training = tf.placeholder(tf.bool, name='is_training')
adnet = ADNetwork()
adnet.create_network(input_node, tensor_lb_action, tensor_lb_class, is_training)
config = tf.ConfigProto(allow_soft_placement=True, log_device_placement=False)
with tf.Session(config=config) as sess:
# load all pretrained weights
adnet.read_original_weights(sess)
variables = tf.trainable_variables()
ps = sess.run(variables)
# zero input
zeros = np.zeros(shape=(1, 2, 107, 107, 3), dtype=np.float32)
from commons import imread
img1 = imread(r'./dataset/OTB/Car4/img/0009.jpg')
from boundingbox import crop_resize
zeros = crop_resize(img1, (70, 51, 103, 87))[None]
action_out, class_out = sess.run([adnet.layer_actions, adnet.layer_scores], feed_dict={input_node: zeros})
print(action_out, class_out)