v-sense / deepnormals Goto Github PK

Hi,
Thank you for open-sourcing the project code!

I am getting the following error when I run

python3 main.py
even though I can cd into the folder just fine

Error: ./Net/DeepNormals/DeepNormals Not Found

Do you have any idea on how I can fix it?

Thanks,

Shahzeb

isn't it 256x256x4, since it concatenated Network (3 channel) and skip blue channel (1 channel)

I cannot find training code

Hi @MatisHudon I am trying to generate DepthNormals for my own datasets. I want to know how to generate mask for sketch.

Hello, thank you for your great job about deepnormals, and I have some doubts about the principle of interactive rendering. Would you share the idea of how to render images from the normal map? Thank you!

Hi. We are the team of style2paints. We are interested in sketch/manga/anime/cartoon researches. I am personally very impressed by this work. I am looking for possible corporation. We have some neat models and many databases. We can offer any possible resources to build up our friendship.

When training, did you train every tile of an image among 420 images or just a random tile among 84000 tiles?

…or is this actually a self-written once?

Just curious…

Hi @MatisHudon I'm trying to follow the direction for running the docker image, but I'm a bit new to docker and I was hoping you could clarify the arguments for the paths. I'm running it like this

sudo nvidia-docker run -v /PycharmProjects/DeepNormals:/var/lib/docker -it matishudon/dockerdeepn python3 /PycharmProjects/DeepNormals/main.py --docker_path /var/lib/docker

But when running it I get the error python3: can't open file '/PycharmProjects/DeepNormals/main.py': [Errno 2] No such file or directory

Could you please help, thank you in advance!

Dear author, if I want to retrain the network, how do I rewrite the code and in what order?

Hello. I am hereby asking for allowance to re-use parts of your License and add it to my modified BSD 3-Clause License:

Parts I want to reuse:
Any other use, in particular any use for commercial purposes, is prohibited. This includes, without limitation, incorporation in a commercial product, use in a commercial service, or production of other artefacts for commercial purposes.

and
You understand and agree that the authors are under no obligation to provide either maintenance services, update services, notices of latent defects, or corrections of defects with regard to the Software. The authors nevertheless reserve the right to update, modify, or discontinue the Software at any time.

Thank you :)

Hi,

I'm wondering why did you thining the input sketch.
Is there any reason that data should have thick line?

Firstly, thank the authors for providing this excellent project!

OS: Win 8.1 x64

In "interactive rendering", the key code for arrow keys doesn't work under Windows OS. So I use the i,j,k,l keys to replace the arrow keys.

original Interactive_Rendering.py:(Line 188)

    if k == 83 : #right arrow
        Xpos += 100
        print('X =' + str(Xpos))
    if k == 81 : #left
        Xpos -= 100
        print('X =' + str(Xpos))
    if k == 82 : #top
        Ypos -= 100
        print('Y =' + str(Ypos))
    if k == 84 :#bottom
        Ypos += 100
        print('Y =' + str(Ypos))

modified Interactive_Rendering.py:(Line 188)

    if k == 108 : #right: l
        Xpos += 100
        print('X =' + str(Xpos))
    if k == 106 : #left: j
        Xpos -= 100
        print('X =' + str(Xpos))
    if k == 105 : #top: i
        Ypos -= 100
        print('Y =' + str(Ypos))
    if k == 107 :#bottom: k
        Ypos += 100
        print('Y =' + str(Ypos))

Firstly, thank the authors for providing this excellent project!
Here I want to report a small problem and corresponding solution.

I prefer pip to docker. pip always install the latest version of each package, and this has caused some compatibility problems (if you install it in a new environment Now).

pip install numpy
pip install tensorflow-gpu
pip install opencv-python
pip install tqdm
pip install tflearn

And then execute:
python main.py

Error:

Traceback (most recent call last):
  File "main.py", line 57, in <module>
    model = GenerateNet()
  File "D:\DeepNormals\model.py", line 134, in
 GenerateNet
    Network = tf.nn.l2_normalize(Network, dim = 3)
  File "D:\VirtualEnvs\deep_normal\lib\site-packages\tensorflow\python\util\dispatch.py",
 line 201, in wrapper
    return target(*args, **kwargs)
TypeError: l2_normalize_v2() got an unexpected keyword argument 'dim'

[Fix]
I suggest specify the version of each package in requirements.txt.

pip install -r requirements.txt

requirements.txt:

numpy==1.15.1
tensorflow-gpu==1.11.0
opencv-python==3.4.3.18
tqdm==4.26.0
tflearn==0.3.2