pytorch-pose-estimation
pytorch-pose-estimation copied to clipboard
Misalignment error
Hi, you resize(bilinear upsample) the output heatmap and paf to original image size, however, the outputs comes from padded images. Since you have changedthe height-width ratio, would it cause misalignment error?
for i, scale in enumerate(multiplier):
img_test = cv2.resize(img_raw, (0, 0), fx=scale, fy=scale, interpolation=cv2.INTER_CUBIC)
img_test_pad, pad = pad_right_down_corner(img_test, param_stride, param_stride)
img_test_pad = np.transpose(np.float32(img_test_pad[:, :, :, np.newaxis]), (3, 2, 0, 1)) / 256 - 0.5
feed = Variable(torch.from_numpy(img_test_pad)).cuda()
output1, output2 = model(feed)
print(output1.size())
print(output2.size())
heatmap = nn.UpsamplingBilinear2d((img_raw.shape[0], img_raw.shape[1])).cuda()(output2)
paf = nn.UpsamplingBilinear2d((img_raw.shape[0], img_raw.shape[1])).cuda()(output1)