example error

[mohamad-hasan-sohan-ajini]

Hi
I run test_image.py and get the following error:

/home/aj/repo/DFace/dface/core/models.py:8: UserWarning: nn.init.xavier_uniform is now deprecated in favor of nn.init.xavier_uniform_.
nn.init.xavier_uniform(m.weight.data)
/home/aj/repo/DFace/dface/core/models.py:9: UserWarning: nn.init.constant is now deprecated in favor of nn.init.constant_.
nn.init.constant(m.bias, 0.1)
Traceback (most recent call last):
File "test_image.py", line 18, in
bboxs, landmarks = mtcnn_detector.detect_face(img)
File "/home/aj/repo/DFace/dface/core/detect.py", line 622, in detect_face
boxes, boxes_align = self.detect_rnet(img, boxes_align)
File "/home/aj/repo/DFace/dface/core/detect.py", line 391, in detect_rnet
tmp = np.zeros((tmph[i], tmpw[i], 3), dtype=np.uint8)
ValueError: negative dimensions are not allowed

I use pytorch .4 and python 3.5

[kuaikuaikim]

@Hajix please use python2.7 and the pytorch version less than 0.3, i did not test for python3 before.

[thnkim]

@kuaikuaikim Hello, I have the same issue although I tried it using python 2.7.15 and pytorch 0.2.0+0b92e5c.

[rohith14]

@kuaikuaikim and @Hajix : I am running into the same issue. Any update? Thanks

[hanson-young]

Recently, I also encountered the same problem. I compared it with the tensorflow version of mtnnn.I found that there is a problem in the preprocessing of the network. The image needs to be multiplied by 1/255, but Dface is not. When you test, go to view and confirm the input image tensor @kuaikuaikim @Hajix

[AngelaDevHao]

Recently, I also encountered the same problem. I compared it with the tensorflow version of mtnnn.I found that there is a problem in the preprocessing of the network. The image needs to be multiplied by 1/255, but Dface is not. When you test, go to view and confirm the input image tensor @kuaikuaikim @Hajix

Did you run the demo successfully? I found the reg output by PNet are too large, which results in this error. I try to add the 1/255 preprocess, but it didn't work. Do you have some advice? @kuaikuaikim @hanson-young

[leadcain]

remove python2.7 pickle

redefine pickle as below:
import six; from six.moves import cPickle as pickle

and modify the code as below:
with open(det_boxs_file, "rb") as file:
det_boxes = pickle.Unpickler(file).load()