def get_scores_new(self, pos_x, pos_y, scaled_search_area, template_z,
filename, design, final_score_sz):
image = Image.open(filename).convert('RGB')
# avg_chan = ImageStat.Stat(image).mean
# frame_padded_x, npad_x = pad_frame(image, image.size, pos_x, pos_y, scaled_search_area[2], avg_chan)
# x_crops = extract_crops_x(frame_padded_x, npad_x, pos_x, pos_y, scaled_search_area[0], scaled_search_area[1], scaled_search_area[2], design.search_sz)
txs = []
for scale in scaled_search_area:
x = gen_xz(image, Rectangle(pos_x, pos_y, scale, scale), to='x')
tx = Image_to_Tensor(x).unsqueeze(0)
txs.append(tx.squeeze(0))
x_crops = torch.stack(txs)
# print Rectangle(pos_x, pos_y,scale,scale)
template_x = self.branch(Variable(x_crops).cuda())
template_z = template_z.repeat(template_x.size(0), 1, 1, 1)
scores = self.model.head(template_z, template_x)
# scores = self.bn_adjust(scores)
# TODO: any elegant alternator?
scores = scores.squeeze().permute(1, 2, 0).data.cpu().numpy()
scores_up = cv2.resize(scores, (final_score_sz, final_score_sz),
interpolation=cv2.INTER_CUBIC)
scores_up = scores_up.transpose((2, 0, 1))
return image, scores_up
def get_template_z_new(self, pos_x, pos_y, z_sz, image, design):
if isinstance(image, six.string_types):
image = Image.open(image).convert('RGB')
# avg_chan = ImageStat.Stat(image).mean
# frame_padded_z, npad_z = pad_frame(image, image.size, pos_x, pos_y, z_sz, avg_chan)
# z_crops = extract_crops_z(frame_padded_z, npad_z, pos_x, pos_y, z_sz, design.exemplar_sz)
z = gen_xz(image, Rectangle(pos_x, pos_y, z_sz, z_sz), to='z')
# cv2.imshow('z', np.array(z))
tz = Image_to_Tensor(z).unsqueeze(0)
template_z = self.branch(Variable(tz).cuda())
# print template_z same
return image, template_z