Implementation problems about posecnn HOT 10 CLOSED

1. These are model points included in the YCB_Video dataset. I have added these files to the repo. You can do git pull to get them.
2. I need to see how to trim.
3. Yes, the cython includes some other packages needed
4. If you do not install it, you need to comment out several lines in the code that load the synthesize library

from posecnn.

Hi Yu Xiang,

Thanks for your reply and for answering my queries. As I was trying to run the demo version, I get this error.

Traceback (most recent call last):
File "./tools/test_net.py", line 103, in
from networks.factory import get_network
File "/home/rajatmittal/Documents/posecnn/tools/../lib/networks/init.py", line 8, in
from .vgg16 import vgg16
File "/home/rajatmittal/Documents/posecnn/tools/../lib/networks/vgg16.py", line 2, in
from networks.network import Network
File "/home/rajatmittal/Documents/posecnn/tools/../lib/networks/network.py", line 6, in
import backprojecting_layer.backprojecting_op as backproject_op
File "/home/rajatmittal/Documents/posecnn/tools/../lib/backprojecting_layer/backprojecting_op.py", line 5, in
_backprojecting_module = tf.load_op_library(filename)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/load_library.py", line 56, in load_op_library
lib_handle = py_tf.TF_LoadLibrary(library_filename, status)
File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/errors_impl.py", line 473, in exit
c_api.TF_GetCode(self.status.status))
tensorflow.python.framework.errors_impl.NotFoundError: /home/rajatmittal/Documents/posecnn/tools/../lib/backprojecting_layer/backprojecting.so: undefined symbol: ZN10tensorflow7strings6StrCatB5cxx11ERKNS0_8AlphaNumES3

I have been trying to search online but haven't really found any solutions. Do you know what might be the issue. I will keep searching and let you know if I find one.

Python version: python 2.7
Ubuntu version: 16.04
Tensorflow version: 1.4.1
CUDA version: 8.0

Thanks.

from posecnn.

The error may because the flag -D_GLIBCXX_USE_CXX11_ABI=0 in compiling the new layers.
If your tensorflow is compiled with -D_GLIBCXX_USE_CXX11_ABI=0, you need to add -D_GLIBCXX_USE_CXX11_ABI=0 to the make.sh file to compile. I think for pre-built tensorflow, you need to add this flag.

from posecnn.

I added this before, i still error
'nvcc -std=c++11 -c -o average_distance_loss_op_gpu.cu.o average_distance_loss_op_gpu.cu.cc
-I $TF_INC -I$TF_INC/external/nsync/public -D GOOGLE_CUDA=1 -x cu -Xcompiler -fPIC -arch=sm_50

g++ -std=c++11 -shared -o average_distance_loss.so average_distance_loss_op.cc
average_distance_loss_op_gpu.cu.o -I $TF_INC -fPIC -D_GLIBCXX_USE_CXX11_ABI=0 -I$TF_INC/external/nsync/public -fPIC -lcudart -L$CUDA_PATH/lib64 -L$TF_LIB -ltensorflow_framework '

should i add like this?

from posecnn.

running :~/Downloads/PoseCNN-master$ ./experiments/scripts/demo.sh $GPU_ID
I get
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/002_master_chef_can/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/003_cracker_box/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/004_sugar_box/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/005_tomato_soup_can/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/006_mustard_bottle/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/007_tuna_fish_can/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/008_pudding_box/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/009_gelatin_box/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/010_potted_meat_can/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/011_banana/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/019_pitcher_base/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/021_bleach_cleanser/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/024_bowl/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/025_mug/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/035_power_drill/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/036_wood_block/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/037_scissors/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/040_large_marker/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/051_large_clamp/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/052_extra_large_clamp/points.xyz
/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/datasets/../../data/LOV/models/061_foam_brick/points.xyz
data/demo_images/000001-color.png
data/demo_images/000001-depth.png
data/demo_images/000002-color.png
data/demo_images/000002-depth.png
data/demo_images/000003-color.png
data/demo_images/000003-depth.png
data/demo_images/000004-color.png
data/demo_images/000004-depth.png
data/demo_images/000005-color.png
data/demo_images/000005-depth.png
{'intrinsic_matrix': array([[1.066778e+03, 0.000000e+00, 3.129869e+02],
[0.000000e+00, 1.067487e+03, 2.413109e+02],
[0.000000e+00, 0.000000e+00, 1.000000e+00]]), 'factor_depth': 10000.0}
/gpu:0
Traceback (most recent call last):
File "./tools/demo.py", line 118, in
from networks.factory import get_network
File "/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/networks/init.py", line 8, in
from .vgg16 import vgg16
File "/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/networks/vgg16.py", line 2, in
from networks.network import Network
File "/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/networks/network.py", line 6, in
import backprojecting_layer.backprojecting_op as backproject_op
File "/home/nivedithak/Downloads/PoseCNN-master/tools/../lib/backprojecting_layer/backprojecting_op.py", line 5, in
_backprojecting_module = tf.load_op_library(filename)
File "/home/nivedithak/tensorflow2/local/lib/python2.7/site-packages/tensorflow/python/framework/load_library.py", line 56, in load_op_library
lib_handle = py_tf.TF_LoadLibrary(library_filename)
tensorflow.python.framework.errors_impl.NotFoundError: /home/nivedithak/Downloads/PoseCNN-master/tools/../lib/backprojecting_layer/backprojecting.so: undefined symbol: ZN10tensorflow7strings6StrCatB5cxx11ERKNS0_8AlphaNumES3

from posecnn.

I get the same error as above, any idea what tf version we should use?

from posecnn.

Any idea what the fix for this is?

from posecnn.

@iretiayo I made an issue how to build this? There is a fork with detailed instruction that should help you to get this project running

from posecnn.

@nkalavak I have the exact same problem, do you have fixed it?

from posecnn.

@Kaju-Bubanja Kaju-Bubanja, thank u so much, i am not trying your version, but really really happy to find this !!!

from posecnn.