Cascaded-fcn: model test problem

Is there any trick I have neglected?

The result look strange. Make sure you can run the notebook as-is and get correct results, before you make modifications.

The code is same as what you show in notebook, thus I cannot find where the code is wrong, can you give me some guidance? thank you

I met the same problem with you, did you find out? I would appriciate if you can share your way out. @manutdzou

I think the released model is wrong, when I self train my own model and use the code above it works well, and the result is good

@RenieWell @mohamed-ezz

Thats great news @manutdzou . You are more than welcome to write a pull request and offer your trained model to the public. Just upload your model to a public filehoster and modify the readme with the link and your name.

Wow I got the same strange result as your first result. Then I'm sure this released model is not so good. Anyway I rebuild U-Net on TensorFlow, my prediction result is not so good but not strange.

@manutdzou . Hi guys, can you share your code?
Thank you very much.

Hey Everyone,
i just updated the Readme and added a docker images, which runs our code smoothly.
Please have a look in the Readme for more details how to start the docker image.
The expected result should look like this print out.
Best wishes,
Patrick
cascaded_unet_inference.pdf.pdf

@PatrickChrist Hi Patrick, thanks for the great work, but when I try to use the pretrained model, I find that the nvidia-docker is hard to install and could you please share a correct pretrained model without using nvidia-docker

@zakizhou I think because this is a reproducibility issue, Docker is our best bet to achieve that.

nvidia-docker is needed only if you want to process the files on the GPU. You can, however, just use docker if you're ok with running on CPU.

If you're running on linux distro, what are the issues you're facing to install nvidia-docker ?

The models are also shared in https://github.com/IBBM/Cascaded-FCN/tree/master/models/cascadedfcn , you can use them in your host environment (without Docker)

@mohamed-ezz thanks for your reply, I am using ubuntu with no gpus, indeed I have tried docker instead of nvidia-docker but sadly when I tried to import pretrained caffe model, the core of jupyter notebook dumped and I don't understand why. Like what @manutdzou said in this issue, the pretrained model here https://github.com/IBBM/Cascaded-FCN/tree/master/models/cascadedfcn performs badly on the sample image. I installed caffe with conda, do you think it's the wrong version of caffe that caused this problem?

Yes it's likely the caffe version. Please use the docker image.

On Jul 4, 2017 5:45 PM, "Jie Zhou" notifications@github.com wrote:

@mohamed-ezz https://github.com/mohamed-ezz thanks for your reply, I am
using ubuntu with no gpus, indeed I have tried docker instead of
nvidia-docker but sadly when I tried to import pretrained caffe model,
the core of jupyter notebook dumped and I don't understand why. Like what
@manutdzou https://github.com/manutdzou said in this issue, the
pretrained model here https://github.com/IBBM/Cascaded-FCN/tree/master/
models/cascadedfcn performs badly on the sample image. I installed caffe
with conda, do you think it's the wrong version of caffe that caused this
problem?

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
https://github.com/IBBM/Cascaded-FCN/issues/13#issuecomment-312908109,
or mute the thread
https://github.com/notifications/unsubscribe-auth/ADqENQMhAJvhyvV1SZ5ApQelMqNhBlTMks5sKl39gaJpZM4M4ckQ
.

@mohamed-ezz OK, I'd try the model on a server with gpu, thanks again!

No need for a gpu, just use docker with the image in the README.md, instead
of nvidia-docker.

On Jul 4, 2017 6:59 PM, "Jie Zhou" notifications@github.com wrote:

@mohamed-ezz https://github.com/mohamed-ezz OK, I'd try the model on a
server with gpu, thanks again!

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
https://github.com/IBBM/Cascaded-FCN/issues/13#issuecomment-312919887,
or mute the thread
https://github.com/notifications/unsubscribe-auth/ADqENQh6pIiGyQYTbKSOx_2reMDlja3Cks5sKm-GgaJpZM4M4ckQ
.

I have released a version of right liver and lesion model in Baidu can use this model like this

`import sys,os
sys.path.insert(0, '/home/zhou/zou/caffe_ws/python')
sys.path.insert(0,'/home/zhou/zou/Cascaded-FCN/lib')
import numpy as np
from matplotlib import pyplot as plt
import caffe

result_path = "/home/zhou/zou/Cascaded-FCN/code/result/"
if not os.path.exists(result_path):
os.makedirs(result_path)

im_list = open('test_lesion_list.txt', 'r').read().splitlines()

caffe.set_mode_gpu()
caffe.set_device(0)
net_liver = caffe.Net('deploy.prototxt', 'liver.caffemodel', caffe.TEST)
net_lesion = caffe.Net('deploy.prototxt', 'lesion.caffemodel', caffe.TEST)

liver = 1
lesion = 2
for i in range(0,len(im_list)):
im = np.load(im_list[i].split(' ')[0])
mask = np.load(im_list[i].split(' ')[1])
in_ = np.array(im, dtype=np.float32)
in_expand = in_[np.newaxis, ...]
blob = in_expand[np.newaxis, :, :, :]

net_liver.blobs['data'].reshape(*blob.shape)
net_liver.blobs['data'].data[...] = blob
net_liver.forward()
output_liver = net_liver.blobs['prob'].data[0].argmax(axis=0)

net_lesion.blobs['data'].reshape(*blob.shape)
net_lesion.blobs['data'].data[...] = blob
net_lesion.forward()
output_lesion = net_lesion.blobs['prob'].data[0].argmax(axis=0)

output = output_liver
ind_1 = np.where(output_liver ==0)
output_lesion[ind_1] = 255
ind_2 = np.where(output_lesion ==0)
output[ind_2] = 2

plt.figure(figsize=(3*5,10))
plt.subplot(1, 3, 1)
plt.title('CT')
plt.imshow(im[92:-92,92:-92], 'gray')
plt.subplot(1, 3, 2)
plt.title('GT')
plt.imshow(mask, 'gray')
plt.subplot(1, 3, 3)
plt.title('pred')
plt.imshow(output, 'gray')
path = result_path + im_list[i].split(' ')[0].split('/')[-1][0:-3] +'jpg'
plt.savefig(path)
plt.close()

`
some result is shown

@mohamed-ezz @RenieWell @mjiansun @PatrickChrist @PiaoLiangHXD

layer {
name: "data"
type: "Input"
top: "data"
input_param { shape: { dim: 1 dim: 1 dim: 572 dim: 572 } }
}

layer {
name: "conv_d0a-b"
type: "Convolution"
bottom: "data"
top: "d0b"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 64
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}

layer {
name: "relu_d0b"
type: "ReLU"
bottom: "d0b"
top: "d0b"
}
layer {
name: "conv_d0b-c"
type: "Convolution"
bottom: "d0b"
top: "d0c"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 64
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}

layer {
name: "relu_d0c"
type: "ReLU"
bottom: "d0c"
top: "d0c"
}
layer {
name: "pool_d0c-1a"
type: "Pooling"
bottom: "d0c"
top: "d1a"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv_d1a-b"
type: "Convolution"
bottom: "d1a"
top: "d1b"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 128
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}

layer {
name: "relu_d1b"
type: "ReLU"
bottom: "d1b"
top: "d1b"
}
layer {
name: "conv_d1b-c"
type: "Convolution"
bottom: "d1b"
top: "d1c"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 128
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}

layer {
name: "relu_d1c"
type: "ReLU"
bottom: "d1c"
top: "d1c"
}
layer {
name: "pool_d1c-2a"
type: "Pooling"
bottom: "d1c"
top: "d2a"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv_d2a-b"
type: "Convolution"
bottom: "d2a"
top: "d2b"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}

layer {
name: "relu_d2b"
type: "ReLU"
bottom: "d2b"
top: "d2b"
}
layer {
name: "conv_d2b-c"
type: "Convolution"
bottom: "d2b"
top: "d2c"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}

layer {
name: "relu_d2c"
type: "ReLU"
bottom: "d2c"
top: "d2c"
}
layer {
name: "pool_d2c-3a"
type: "Pooling"
bottom: "d2c"
top: "d3a"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv_d3a-b"
type: "Convolution"
bottom: "d3a"
top: "d3b"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}

layer {
name: "relu_d3b"
type: "ReLU"
bottom: "d3b"
top: "d3b"
}
layer {
name: "conv_d3b-c"
type: "Convolution"
bottom: "d3b"
top: "d3c"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}

layer {
name: "relu_d3c"
type: "ReLU"
bottom: "d3c"
top: "d3c"
}

layer {
name: "pool_d3c-4a"
type: "Pooling"
bottom: "d3c"
top: "d4a"
pooling_param {
pool: MAX
kernel_size: 2
stride: 2
}
}
layer {
name: "conv_d4a-b"
type: "Convolution"
bottom: "d4a"
top: "d4b"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 1024
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}

layer {
name: "relu_d4b"
type: "ReLU"
bottom: "d4b"
top: "d4b"
}
layer {
name: "conv_d4b-c"
type: "Convolution"
bottom: "d4b"
top: "d4c"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 1024
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}

layer {
name: "relu_d4c"
type: "ReLU"
bottom: "d4c"
top: "d4c"
}

layer {
name: "upconv_d4c_u3a"
type: "Deconvolution"
bottom: "d4c"
top: "u3a"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 0
kernel_size: 2
stride: 2
weight_filler {
type: "xavier"
}
}
}

layer {
name: "relu_u3a"
type: "ReLU"
bottom: "u3a"
top: "u3a"
}
layer {
name: "crop_d3c-d3cc"
type: "Crop"
bottom: "d3c"
bottom: "u3a"
top: "d3cc"

}
layer {
name: "concat_d3cc_u3a-b"
type: "Concat"
bottom: "u3a"
bottom: "d3cc"
top: "u3b"
}
layer {
name: "conv_u3b-c"
type: "Convolution"
bottom: "u3b"
top: "u3c"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}
layer {
name: "relu_u3c"
type: "ReLU"
bottom: "u3c"
top: "u3c"
}
layer {
name: "conv_u3c-d"
type: "Convolution"
bottom: "u3c"
top: "u3d"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 512
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}
layer {
name: "relu_u3d"
type: "ReLU"
bottom: "u3d"
top: "u3d"
}
layer {
name: "upconv_u3d_u2a"
type: "Deconvolution"
bottom: "u3d"
top: "u2a"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 0
kernel_size: 2
stride: 2
weight_filler {
type: "xavier"
}
}
}
layer {
name: "relu_u2a"
type: "ReLU"
bottom: "u2a"
top: "u2a"
}
layer {
name: "crop_d2c-d2cc"
type: "Crop"
bottom: "d2c"
bottom: "u2a"
top: "d2cc"

}
layer {
name: "concat_d2cc_u2a-b"
type: "Concat"
bottom: "u2a"
bottom: "d2cc"
top: "u2b"
}
layer {
name: "conv_u2b-c"
type: "Convolution"
bottom: "u2b"
top: "u2c"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}
layer {
name: "relu_u2c"
type: "ReLU"
bottom: "u2c"
top: "u2c"
}
layer {
name: "conv_u2c-d"
type: "Convolution"
bottom: "u2c"
top: "u2d"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 256
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}
layer {
name: "relu_u2d"
type: "ReLU"
bottom: "u2d"
top: "u2d"
}
layer {
name: "upconv_u2d_u1a"
type: "Deconvolution"
bottom: "u2d"
top: "u1a"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 128
pad: 0
kernel_size: 2
stride: 2
weight_filler {
type: "xavier"
}
}
}
layer {
name: "relu_u1a"
type: "ReLU"
bottom: "u1a"
top: "u1a"
}
layer {
name: "crop_d1c-d1cc"
type: "Crop"
bottom: "d1c"
bottom: "u1a"
top: "d1cc"

}
layer {
name: "concat_d1cc_u1a-b"
type: "Concat"
bottom: "u1a"
bottom: "d1cc"
top: "u1b"
}
layer {
name: "conv_u1b-c"
type: "Convolution"
bottom: "u1b"
top: "u1c"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 128
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}
layer {
name: "relu_u1c"
type: "ReLU"
bottom: "u1c"
top: "u1c"
}
layer {
name: "conv_u1c-d"
type: "Convolution"
bottom: "u1c"
top: "u1d"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 128
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}
layer {
name: "relu_u1d"
type: "ReLU"
bottom: "u1d"
top: "u1d"
}
layer {
name: "upconv_u1d_u0a_NEW"
type: "Deconvolution"
bottom: "u1d"
top: "u0a"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 64
pad: 0
kernel_size: 2
stride: 2
weight_filler {
type: "xavier"
}
}
}
layer {
name: "relu_u0a"
type: "ReLU"
bottom: "u0a"
top: "u0a"
}
layer {
name: "crop_d0c-d0cc"
type: "Crop"
bottom: "d0c"
bottom: "u0a"
top: "d0cc"

}
layer {
name: "concat_d0cc_u0a-b"
type: "Concat"
bottom: "u0a"
bottom: "d0cc"
top: "u0b"
}
layer {
name: "conv_u0b-c_New"
type: "Convolution"
bottom: "u0b"
top: "u0c"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 64
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}
layer {
name: "relu_u0c"
type: "ReLU"
bottom: "u0c"
top: "u0c"
}
layer {
name: "conv_u0c-d_New"
type: "Convolution"
bottom: "u0c"
top: "u0d"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 64
pad: 0
kernel_size: 3
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}
layer {
name: "relu_u0d"
type: "ReLU"
bottom: "u0d"
top: "u0d"
}
layer {
name: "conv_u0d-score_New"
type: "Convolution"
bottom: "u0d"
top: "score"
param {
lr_mult: 1
decay_mult: 1
}
param {
lr_mult: 2
decay_mult: 0
}
convolution_param {
num_output: 2
pad: 0
kernel_size: 1
weight_filler {
type: "xavier"
}
engine: CAFFE
}
}

layer {
name: "prob"
type: "Softmax"
bottom: "score"
top: "prob"
}

Great work @manutdzou
Thanks for your support. Would you mind to commit your work in this repo?
We could have a folder model-zoo/manutdzou in which you post your code as notebook and your prototxt and the links to baidu as text file? Other users will definitly appreciate. If you have a paper about your work we can also add this.

Ok, if my code and model works well, I will be glad to commit in this repo!

发自网易邮箱大师
On 07/07/2017 17:33, Patrick Christ wrote:

Great work @manutdzou
Thanks for your support. Would you mind to commit your work in this repo?
We could have a folder model-zoo/manutdzou in which you post your code as notebook and your prototxt and the links to baidu as text file? Other users will definitly appreciate. If you have a paper about your work we can also add this.

—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or mute the thread.