As mentioned in the paper: For head, hands and feet, we use the manually obtained UV fields provided in the SMPL model. For the rest of the parts we obtain the unwrapping via multidimensional scaling applied to pairwise geodesic distances. Then we get this：

Is there any code to generate this UV map? And what is the corresponence between this map with 6890 SMPL vertices?

when run the make ops,i encountered this error and i also have found that many people had the same problems i had,but none of them were soved. (base) tensor@tensor-GS65-Stealth-Thin-8RF:~/densepose$ make ops mkdir -p build && cd build && cmake .. && make -j12 CMake Error at CMakeLists.txt:8 (find_package): By not providing "FindCaffe2.cmake" in CMAKE_MODULE_PATH this project has asked CMake to find a package configuration file provided by "Caffe2", but CMake did not find one.

Could not find a package configuration file provided by "Caffe2" with any of the following names:

Caffe2Config.cmake
caffe2-config.cmake

Add the installation prefix of "Caffe2" to CMAKE_PREFIX_PATH or set "Caffe2_DIR" to a directory containing one of the above files. If "Caffe2" provides a separate development package or SDK, be sure it has been installed.

-- Configuring incomplete, errors occurred! See also "/home/tensor/densepose/build/CMakeFiles/CMakeOutput.log". Makefile:13: recipe for target 'ops' failed make: *** [ops] Error 1

this is the last step to use this demo,please give me a hand!

I succeeded densepose test for video with reference to (https://github.com/trrahul/densepose-video). while tests, i got a question.

1. When I detect an object using densepose, where are the keypoints' coordination stored? (which variable?) Example) get_keypoints() in keypoints.py ? 2. What information can I use in densepose data?

Thank you! i want to your detail opinion.

When I run "make ops" I encountered many errors which made me very sad, most of them were solved in #152, thanks to @linkinpark213. However, here is a new error I met which was not emerged in that issue. The error message looks like this:

[ 12%] Building NVCC (Device) object CMakeFiles/caffe2_detectron_custom_ops_gpu.dir/detectron/ops/caffe2_detectron_custom_ops_gpu_generated_zero_even_op.cu.o [ 25%] Building NVCC (Device) object CMakeFiles/caffe2_detectron_custom_ops_gpu.dir/detectron/ops/caffe2_detectron_custom_ops_gpu_generated_pool_points_interp.cu.o [ 37%] Building CXX object CMakeFiles/caffe2_detectron_custom_ops.dir/detectron/ops/pool_points_interp.cc.o [ 50%] Building CXX object CMakeFiles/caffe2_detectron_custom_ops.dir/detectron/ops/zero_even_op.cc.o [ 62%] Linking CXX shared library libcaffe2_detectron_custom_ops.so [ 62%] Built target caffe2_detectron_custom_ops Scanning dependencies of target caffe2_detectron_custom_ops_gpu make[2]: *** No rule to make target /usr/local/cuda/lib64/libculibos.a', needed bylibcaffe2_detectron_custom_ops_gpu.so'. Stop. make[2]: *** Waiting for unfinished jobs.... [ 75%] Building CXX object CMakeFiles/caffe2_detectron_custom_ops_gpu.dir/detectron/ops/pool_points_interp.cc.o [ 87%] Building CXX object CMakeFiles/caffe2_detectron_custom_ops_gpu.dir/detectron/ops/zero_even_op.cc.o make[1]: *** [CMakeFiles/caffe2_detectron_custom_ops_gpu.dir/all] Error 2 make: *** [all] Error 2

The problem is when compiling the library "libcaffe_detectron_custom_ops_gpu.so", the compiler can't find the dependent library "libculibos.a" from the path "/usr/local/cuda/lib64". I found that this path in fact doesn't exist on my system since I didn't create a symbol link for a specific version of CUDA.

By checking the file "CMakeFiles/caffe2_detectron_custom_ops_gpu.dir/build.cmake" generated by "cmake .." command, I found all paths related to cuda libraries were generated correctly with specefic cuda version (e.g., /usr/local/cuda-9.0/lib64/libcublas.so) except only this one (i.e., libculibos.a).

So, here's my solution which I think should solve this problem (but I still don't know why this happened, maybe it's a minor bug):

1. Execute cmake .. in $DENSEPOSE/build;
2. Check $DENSEPOSE/build/CMakeFiles/caffe2_detectron_custom_ops_gpu.dir/build.cmake and $DENSEPOSE/build/CMakeFiles/caffe2_detectron_custom_ops_gpu.dir/link.txt to make sure all dependencies can be found in the specified paths. (In my case, replace /usr/local/cuda/lib64 with /usr/local/cuda-9.0/lib64 in both files);
3. make -j8 (or simply make). After this, you can test your compilation by running $DENSEPOSE/detectron/tests/test_zero_even_op.py

Although this is not a big issue, but there's still a chance that someone may encounter it. So, hope this can help :)

Thanks for making this tremendous work public available! I have read the code of tools/infer_simple.py and the output of is UV coordinate. But I am more interested in the estimate SMPL model and I am wondering is it possible to output the 2D coordinates of the 6890 vertices given an image?

cd $DensPose && make ops when I perform this instruction , it will tell me that ‘ mkdir -p build && cd build && cmake .. && make -j48 -- Caffe2: Cannot find gflags automatically. Using legacy find. -- Caffe2: Found gflags (include: /usr/include, library: /usr/lib/x86_64-linux-gnu/libgflags.so) -- Caffe2: Cannot find glog automatically. Using legacy find. -- Caffe2: Found glog (include: /usr/include, library: /usr/lib/x86_64-linux-gnu/libglog.so) -- Caffe2: Found protobuf with old-style protobuf targets. -- Caffe2: Protobuf version CMake Error at /data/xiaobing.wang/peixuan.liang/caffe2/build/Caffe2Config.cmake:55 (message): Your installed Caffe2 is built with protobuf , while your current cmake setting discovers protobuf version . Please specify a protobuf version that is the same as the built version. Call Stack (most recent call first): CMakeLists.txt:8 (find_package)

-- Configuring incomplete, errors occurred! See also "/data/xiaobing.wang/peixuan.liang/DensePose-master/build/CMakeFiles/CMakeOutput.log". See also "/data/xiaobing.wang/peixuan.liang/DensePose-master/build/CMakeFiles/CMakeError.log". Makefile:13: recipe for target 'ops' failed make: *** [ops] Error 1

’ and i could not solve this problem，what should i do? thanks very much.

I set up caffe2 and Densepose environment per instruction, there's something wrong with making custom ops though, but the rest seems okay. However, the generated results are meaningless.

Here's my command:

python2 tools/infer_simple.py \

	--cfg configs/DensePose_ResNet101_FPN_s1x-e2e.yaml \

	--output-dir DensePoseData/infer_out_new/ \

	--image-ext jpg \

	--wts https://s3.amazonaws.com/densepose/DensePose_ResNet101_FPN_s1x-e2e.pkl \

	DensePoseData/demo_data/demo_im.jpg`

Here is the output info, all the bounding box score has 0.5000 which is below the threshold 0.7. note that I added some print function in the utils/vis.py to force the program save the .pdf result even if return criterion has been triggered:

(py27-caffe2) vcl@vcl-dl-2:/backup/DensePose$ bash run_test.sh

Found Detectron ops lib: /home/vcl/anaconda3/envs/py27-caffe2/lib/libcaffe2_detectron_ops_gpu.so

E0907 14:51:03.959839  6653 init_intrinsics_check.cc:43] CPU feature avx is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.

E0907 14:51:03.959861  6653 init_intrinsics_check.cc:43] CPU feature avx2 is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.

E0907 14:51:03.959867  6653 init_intrinsics_check.cc:43] CPU feature fma is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.

WARNING cnn.py:  25: [====DEPRECATE WARNING====]: you are creating an object from CNNModelHelper class which will be deprecated soon. Please use ModelHelper object with brew module. For more information, please refer to caffe2.ai and python/brew.py, python/brew_test.py for more information.

INFO net.py:  51: Loading weights from: /tmp/detectron-download-cache/DensePose_ResNet101_FPN_s1x-e2e.pkl

I0907 14:51:04.730620  6653 operator.cc:191] Engine CUDNN is not available for operator MaxPool.

I0907 14:51:04.736507  6653 operator.cc:191] Engine CUDNN is not available for operator MaxPool.

I0907 14:51:04.737638  6653 net_dag_utils.cc:102] Operator graph pruning prior to chain compute took: 8.1733e-05 secs

I0907 14:51:04.741101  6653 operator.cc:191] Engine CUDNN is not available for operator MaxPool.

I0907 14:51:04.747277  6653 operator.cc:191] Engine CUDNN is not available for operator MaxPool.

I0907 14:51:04.747470  6653 net_dag_utils.cc:102] Operator graph pruning prior to chain compute took: 7.0287e-05 secs

I0907 14:51:04.749591  6653 net_dag_utils.cc:102] Operator graph pruning prior to chain compute took: 1.0054e-05 secs

INFO infer_simple.py: 103: Processing DensePoseData/demo_data/demo_im.jpg -> DensePoseData/infer_out_new/demo_im.jpg.pdf

I0907 14:51:07.475107  6653 net_async_base.cc:435] Using specified CPU pool size: 4; NUMA node id: -1

I0907 14:51:07.475139  6653 net_async_base.cc:440] Created new CPU pool, size: 4; NUMA node id: -1

I0907 14:51:08.376976  6653 net_async_base.cc:435] Using specified CPU pool size: 4; NUMA node id: -1

I0907 14:51:08.377008  6653 net_async_base.cc:440] Created new CPU pool, size: 4; NUMA node id: -1

INFO infer_simple.py: 115: Inference time: 3.384s

INFO infer_simple.py: 117:  | im_detect_bbox: 0.935s

INFO infer_simple.py: 117:  | im_detect_body_uv: 2.445s

INFO infer_simple.py: 117:  | misc_bbox: 0.003s

INFO infer_simple.py: 120:  \ Note: inference on the first image will be slower than the rest (caches and auto-tuning need to warm up)

going in vis

boxes: (395, 5)

[[0.0000000e+00 0.0000000e+00 1.1634885e+02 1.5376009e+02 5.0000000e-01]
 [8.7234839e+02 0.0000000e+00 1.0807764e+03 1.2694235e+02 5.0000000e-01]
 [4.1836800e+01 0.0000000e+00 2.5026465e+02 1.2694235e+02 5.0000000e-01]
 ...
 [9.6155505e+02 0.0000000e+00 1.0129530e+03 3.1035259e+01 5.0000000e-01]
 [1.0553225e+03 0.0000000e+00 1.1067205e+03 3.1035259e+01 5.0000000e-01]
 [9.3476440e+02 0.0000000e+00 9.8616235e+02 3.1035259e+01 5.0000000e-01]]

DensePoseData/demo_data/demo_im.jpg
IUV written to:  DensePoseData/infer_out_new/demo_im_IUV.png

And here's the inferred results(pdf and converted .png): demo_im.jpg.pdf demo_im_jpg

I didn't change anything in the infer script (except some debugging print functions), and I don't know where goes wrong. Is it possible for you to check it out? Thanks!

Environment ubuntu 18.04 python=2.7 caffe2=0.8.dev (installed from anaconda) cuda=9.0.176 cudnn=7.2.1

Hi,

First of all, thank you for making such a excellent work public available.

I downloaded the uv data and found the number of faces in the data is 13774. But there are 13776 faces in the SMPL mesh, which has two more faces.

I wonder if there is anything I missed? How to make a correspondence between the uv coordinates and SMPL mesh vertices?

Thanks a lot!

I installed Caffe2 using the Anaconda pre-built binary with CUDA 9.0 and cuDNN 7.1.4. infer_simple.py seems to work just fine without any errors, but I do not think that it is making use of the GPU, and it outputs warnings like:

Engine CUDNN is not available for operator MaxPool

Also, there have been new messages like Created new CPU Pool size: 4; NUMA node id: -1, but I don't think it should be using the CPU since I have the GPU. Does anyone know what's wrong?

I am under Ubuntu 16.04, Cuda 8.0 CuDNN 6.0.21; on a GTX 980Ti. I attach my caffe2 cmake summary and my compile log for ref compile_pytorch.txt cmake_pytorch.txt

Cannot figure out where is the problem. HELP please!!!

I pass the first caffe2 tests:

1. python -c 'from caffe2.python import core' 2>/dev/null && echo "Success" || echo "Failure" Success
2. python2 -c 'from caffe2.python import workspace; print(workspace.NumCudaDevices())' 1

Then I "fail" the relu_op_test.py with a weird message but I have read those errors may just be warnings so not sure.

 python caffe2/python/operator_test/relu_op_test.py
No handlers could be found for logger "caffe2.python.net_drawer"
net_drawer will not run correctly. Please install the correct dependencies.
/usr/local/lib/python2.7/dist-packages/caffe2/python/hypothesis_test_util.py:75: HypothesisDeprecationWarning: 
The min_satisfying_examples setting has been deprecated and disabled, due to
overlap with the filter_too_much healthcheck and poor interaction with the
max_examples setting.

  verbosity=hypothesis.Verbosity.verbose))
/usr/local/lib/python2.7/dist-packages/caffe2/python/hypothesis_test_util.py:84: HypothesisDeprecationWarning: 
The min_satisfying_examples setting has been deprecated and disabled, due to
overlap with the filter_too_much healthcheck and poor interaction with the
max_examples setting.

  verbosity=hypothesis.Verbosity.verbose))
/usr/local/lib/python2.7/dist-packages/caffe2/python/hypothesis_test_util.py:92: HypothesisDeprecationWarning: 
The min_satisfying_examples setting has been deprecated and disabled, due to
overlap with the filter_too_much healthcheck and poor interaction with the
max_examples setting.

  verbosity=hypothesis.Verbosity.verbose))
E0702 22:04:28.531708  3503 init_intrinsics_check.cc:43] CPU feature avx is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.
E0702 22:04:28.531721  3503 init_intrinsics_check.cc:43] CPU feature avx2 is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.
E0702 22:04:28.531724  3503 init_intrinsics_check.cc:43] CPU feature fma is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.
Trying example: test_relu(self=<__main__.TestRelu testMethod=test_relu>, X=array([0.], dtype=float32), gc=, dc=[, device_type: 1], engine=u'')
{u'X': }
{u'X': device_type: 1
}
Trying example: test_relu(self=<__main__.TestRelu testMethod=test_relu>, X=array([[[[-0.25521252],
         [-0.25521252],
         [-0.25521252],
         [-0.25521252]],

        [[-0.25521252],
         [-0.25521252],
         [-0.25521252],
         [-0.25521252]],

        [[-0.25521252],
         [-0.25521252],
         [-0.25521252],
         [-0.25521252]]],


       [[[-0.25521252],
         [-0.25521252],
         [ 0.37402993],
         [-0.25521252]],

        [[-0.25521252],
         [-0.25521252],
         [-0.25521252],
         [-0.25521252]],

        [[-0.25521252],
         [-0.25521252],
         [-0.25521252],
         [-0.25521252]]],


       [[[-0.25521252],
         [-0.25521252],
         [ 0.6883273 ],
         [-0.25521252]],

        [[-0.25521252],
         [-0.25521252],
         [-0.25521252],
         [-0.25521252]],

        [[-0.25521252],
         [ 0.86089224],
         [-0.25521252],
         [-0.25521252]]]], dtype=float32), gc=device_type: 1, dc=[, device_type: 1], engine=u'')
{u'X': }
{u'X': device_type: 1
}
Trying example: test_relu(self=<__main__.TestRelu testMethod=test_relu>, X=array([[-0.793503  , -0.793503  , -0.793503  ],
       [-0.49327314, -0.793503  , -0.793503  ],
       [-0.793503  , -0.793503  , -0.793503  ],
       [-0.29144055, -0.793503  , -0.793503  ],
       [-0.793503  ,  0.07806086, -0.5510364 ]], dtype=float32), gc=device_type: 1, dc=[, device_type: 1], engine=u'CUDNN')
{u'X': }
I0702 22:04:28.761914  3503 operator.cc:169] Engine CUDNN is not available for operator Relu.
{u'X': device_type: 1
}
Traceback (most recent call last):
  File "/usr/local/lib/python2.7/dist-packages/hypothesis/core.py", line 671, in evaluate_test_data
    result = self.execute(data, collect=True)
  File "/usr/local/lib/python2.7/dist-packages/hypothesis/core.py", line 586, in execute
    result = self.test_runner(data, run)
  File "/usr/local/lib/python2.7/dist-packages/hypothesis/executors.py", line 58, in default_new_style_executor
    return function(data)
  File "/usr/local/lib/python2.7/dist-packages/hypothesis/core.py", line 582, in run
    return test(*args, **kwargs)
  File "caffe2/python/operator_test/relu_op_test.py", line 19, in test_relu
    engine=st.sampled_from(["", "CUDNN"]),
  File "/usr/local/lib/python2.7/dist-packages/hypothesis/core.py", line 524, in test
    result = self.test(*args, **kwargs)
  File "caffe2/python/operator_test/relu_op_test.py", line 26, in test_relu
    self.assertDeviceChecks(dc, op, [X], [0])
  File "/usr/local/lib/python2.7/dist-packages/caffe2/python/hypothesis_test_util.py", line 350, in assertDeviceChecks
    dc.CheckSimple(op, inputs, outputs_to_check, input_device_options)
  File "/usr/local/lib/python2.7/dist-packages/caffe2/python/device_checker.py", line 49, in CheckSimple
    workspace.RunOperatorOnce(op)
  File "/usr/local/lib/python2.7/dist-packages/caffe2/python/workspace.py", line 165, in RunOperatorOnce
    return C.run_operator_once(StringifyProto(operator))
RuntimeError: [enforce fail at context_gpu.h:100] status == CUDNN_STATUS_SUCCESS. 4 vs 0. , Error at: /home/tets/pytorch/caffe2/core/context_gpu.h:100: CUDNN_STATUS_INTERNAL_ERROR Error from operator: 
input: "X" output: "Y" name: "" type: "Relu" device_option { device_type: 1 } engine: "CUDNN"

I compiled denspose with no pb, and passed the detectron checks:

python2 $DENSEPOSE/detectron/tests/test_spatial_narrow_as_op.py
No handlers could be found for logger "caffe2.python.net_drawer"
net_drawer will not run correctly. Please install the correct dependencies.
E0702 22:22:49.913838  3731 init_intrinsics_check.cc:43] CPU feature avx is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.
E0702 22:22:49.913887  3731 init_intrinsics_check.cc:43] CPU feature avx2 is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.
E0702 22:22:49.913918  3731 init_intrinsics_check.cc:43] CPU feature fma is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.
Found Detectron ops lib: /usr/local/lib/libcaffe2_detectron_ops_gpu.so
...
----------------------------------------------------------------------
Ran 3 tests in 0.932s

OK

 python2 $DENSEPOSE/detectron/tests/test_zero_even_op.py
E0702 22:24:07.535202  3750 init_intrinsics_check.cc:43] CPU feature avx is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.
E0702 22:24:07.535217  3750 init_intrinsics_check.cc:43] CPU feature avx2 is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.
E0702 22:24:07.535226  3750 init_intrinsics_check.cc:43] CPU feature fma is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.
............
----------------------------------------------------------------------
Ran 12 tests in 0.136s

OK

But when running the infer_simple.py command, I got the below cryptic error message.

~/densepose$ python2 tools/infer_simple.py     --cfg configs/DensePose_ResNet101_FPN_s1x-e2e.yaml     --output-dir DensePoseData/infer_out/     --image-ext jpg     --wts https://s3.amazonaws.com/densepose/DensePose_ResNet101_FPN_s1x-e2e.pkl     DensePoseData/demo_data/20170716_101313.jpg
/home/tets/.local/lib/python2.7/site-packages/h5py/__init__.py:34: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.
  from ._conv import register_converters as _register_converters
Found Detectron ops lib: /usr/local/lib/libcaffe2_detectron_ops_gpu.so
E0702 21:59:20.453374  3391 init_intrinsics_check.cc:43] CPU feature avx is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.
E0702 21:59:20.453388  3391 init_intrinsics_check.cc:43] CPU feature avx2 is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.
E0702 21:59:20.453392  3391 init_intrinsics_check.cc:43] CPU feature fma is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU.
WARNING cnn.py:  25: [====DEPRECATE WARNING====]: you are creating an object from CNNModelHelper class which will be deprecated soon. Please use ModelHelper object with brew module. For more information, please refer to caffe2.ai and python/brew.py, python/brew_test.py for more information.
INFO net.py:  51: Loading weights from: /tmp/detectron-download-cache/DensePose_ResNet101_FPN_s1x-e2e.pkl
I0702 21:59:21.176241  3391 net_dag_utils.cc:102] Operator graph pruning prior to chain compute took: 0.000106374 secs
I0702 21:59:21.185215  3391 net_dag_utils.cc:102] Operator graph pruning prior to chain compute took: 8.9347e-05 secs
I0702 21:59:21.186816  3391 net_dag_utils.cc:102] Operator graph pruning prior to chain compute took: 1.1866e-05 secs
INFO infer_simple.py: 103: Processing DensePoseData/demo_data/20170716_101313.jpg -> DensePoseData/infer_out/20170716_101313.jpg.pdf
I0702 21:59:21.589475  3391 net_async_base.cc:435] Using specified CPU pool size: 4; NUMA node id: -1
I0702 21:59:21.589490  3391 net_async_base.cc:440] Created new CPU pool, size: 4; NUMA node id: -1
E0702 21:59:23.285727  3413 net_async_base.cc:368] [enforce fail at context_gpu.h:100] status == CUDNN_STATUS_SUCCESS. 4 vs 0. , Error at: /home/tets/pytorch/caffe2/core/context_gpu.h:100: CUDNN_STATUS_INTERNAL_ERROR Error from operator: 
input: "gpu_0/data" input: "gpu_0/conv1_w" output: "gpu_0/conv1" name: "" type: "Conv" arg { name: "kernel" i: 7 } arg { name: "exhaustive_search" i: 0 } arg { name: "pad" i: 3 } arg { name: "order" s: "NCHW" } arg { name: "stride" i: 2 } device_option { device_type: 1 cuda_gpu_id: 0 } engine: "CUDNN",  op Conv
WARNING workspace.py: 185: Original python traceback for operator `0` in network `generalized_rcnn` in exception above (most recent call last):
WARNING workspace.py: 190:   File "tools/infer_simple.py", line 140, in <module>
WARNING workspace.py: 190:   File "tools/infer_simple.py", line 91, in main
WARNING workspace.py: 190:   File "/home/tets/densepose/detectron/core/test_engine.py", line 334, in initialize_model_from_cfg
WARNING workspace.py: 190:   File "/home/tets/densepose/detectron/modeling/model_builder.py", line 119, in create
WARNING workspace.py: 190:   File "/home/tets/densepose/detectron/modeling/model_builder.py", line 84, in generalized_rcnn
WARNING workspace.py: 190:   File "/home/tets/densepose/detectron/modeling/model_builder.py", line 233, in build_generic_detection_model
WARNING workspace.py: 190:   File "/home/tets/densepose/detectron/modeling/optimizer.py", line 46, in build_data_parallel_model
WARNING workspace.py: 190:   File "/home/tets/densepose/detectron/modeling/model_builder.py", line 165, in _single_gpu_build_func
WARNING workspace.py: 190:   File "/home/tets/densepose/detectron/modeling/FPN.py", line 55, in add_fpn_ResNet101_conv5_body
WARNING workspace.py: 190:   File "/home/tets/densepose/detectron/modeling/FPN.py", line 96, in add_fpn_onto_conv_body
WARNING workspace.py: 190:   File "/home/tets/densepose/detectron/modeling/ResNet.py", line 40, in add_ResNet101_conv5_body
WARNING workspace.py: 190:   File "/home/tets/densepose/detectron/modeling/ResNet.py", line 90, in add_ResNet_convX_body
WARNING workspace.py: 190:   File "/home/tets/densepose/detectron/modeling/ResNet.py", line 243, in basic_bn_stem
WARNING workspace.py: 190:   File "/usr/local/lib/python2.7/dist-packages/caffe2/python/cnn.py", line 97, in Conv
WARNING workspace.py: 190:   File "/usr/local/lib/python2.7/dist-packages/caffe2/python/brew.py", line 107, in scope_wrapper
WARNING workspace.py: 190:   File "/usr/local/lib/python2.7/dist-packages/caffe2/python/helpers/conv.py", line 186, in conv
WARNING workspace.py: 190:   File "/usr/local/lib/python2.7/dist-packages/caffe2/python/helpers/conv.py", line 139, in _ConvBase
Traceback (most recent call last):
  File "tools/infer_simple.py", line 140, in <module>
    main(args)
  File "tools/infer_simple.py", line 109, in main
    model, im, None, timers=timers
  File "/home/tets/densepose/detectron/core/test.py", line 58, in im_detect_all
    model, im, cfg.TEST.SCALE, cfg.TEST.MAX_SIZE, boxes=box_proposals
  File "/home/tets/densepose/detectron/core/test.py", line 158, in im_detect_bbox
    workspace.RunNet(model.net.Proto().name)
  File "/usr/local/lib/python2.7/dist-packages/caffe2/python/workspace.py", line 217, in RunNet
    StringifyNetName(name), num_iter, allow_fail,
  File "/usr/local/lib/python2.7/dist-packages/caffe2/python/workspace.py", line 178, in CallWithExceptionIntercept
    return func(*args, **kwargs)
RuntimeError: [enforce fail at context_gpu.h:100] status == CUDNN_STATUS_SUCCESS. 4 vs 0. , Error at: /home/tets/pytorch/caffe2/core/context_gpu.h:100: CUDNN_STATUS_INTERNAL_ERROR Error from operator: 
input: "gpu_0/data" input: "gpu_0/conv1_w" output: "gpu_0/conv1" name: "" type: "Conv" arg { name: "kernel" i: 7 } arg { name: "exhaustive_search" i: 0 } arg { name: "pad" i: 3 } arg { name: "order" s: "NCHW" } arg { name: "stride" i: 2 } device_option { device_type: 1 cuda_gpu_id: 0 } engine: "CUDNN"

my NVIDIA setup for ref:

Mon Jul  2 22:20:18 2018       
+-----------------------------------------------------------------------------+
| NVIDIA-SMI 384.130                Driver Version: 384.130                   |
|-------------------------------+----------------------+----------------------+
| GPU  Name        Persistence-M| Bus-Id        Disp.A | Volatile Uncorr. ECC |
| Fan  Temp  Perf  Pwr:Usage/Cap|         Memory-Usage | GPU-Util  Compute M. |
|===============================+======================+======================|
|   0  GeForce GTX 980 Ti  Off  | 00000000:01:00.0  On |                  N/A |
|  0%   50C    P0    61W / 250W |    243MiB /  6077MiB |      0%      Default |
+-------------------------------+----------------------+----------------------+
                                                                               
+-----------------------------------------------------------------------------+
| Processes:                                                       GPU Memory |
|  GPU       PID   Type   Process name                             Usage      |
|=============================================================================|
|    0      1141      G   /usr/lib/xorg/Xorg                            17MiB |
|    0      1447      G   /usr/lib/xorg/Xorg                           101MiB |
|    0      1917      G   /usr/bin/gnome-shell                          93MiB |
+-----------------------------------------------------------------------------+

When I run

python2 tools/infer_simple.py \
    --cfg configs/DensePose_ResNet101_FPN_s1x-e2e.yaml \
    --output-dir DensePoseData/infer_out/ \
    --image-ext jpg \
    --wts https://s3.amazonaws.com/densepose/DensePose_ResNet101_FPN_s1x-e2e.pkl \
    DensePoseData/demo_data/demo_im.jpg

from the Getting Started page, I get the following problem. Doesn't look like the yaml parser implemented the proper constructors. I built stuff from source and I verified that all the tests in the install guide passes. Anyone else have this issue? Thanks.

Traceback (most recent call last):
  File "tools/infer_simple.py", line 140, in <module>
    main(args)
  File "tools/infer_simple.py", line 91, in main
    model = infer_engine.initialize_model_from_cfg(args.weights)
  File "/home/ubuntu/densepose/detectron/core/test_engine.py", line 333, in initialize_model_from_cfg
    model, weights_file, gpu_id=gpu_id,
  File "/home/ubuntu/densepose/detectron/utils/net.py", line 56, in initialize_gpu_from_weights_file
    saved_cfg = load_cfg(src_blobs['cfg'])
  File "/home/ubuntu/densepose/detectron/core/config.py", line 1164, in load_cfg
    return yaml.load(cfg_to_load)
  File "/home/ubuntu/.local/lib/python2.7/site-packages/yaml/__init__.py", line 74, in load
    return loader.get_single_data()
  File "/home/ubuntu/.local/lib/python2.7/site-packages/yaml/constructor.py", line 40, in get_single_data
    return self.construct_document(node)
  File "/home/ubuntu/.local/lib/python2.7/site-packages/yaml/constructor.py", line 44, in construct_document
    data = self.construct_object(node)
  File "/home/ubuntu/.local/lib/python2.7/site-packages/yaml/constructor.py", line 89, in construct_object
    data = constructor(self, node)
  File "/home/ubuntu/.local/lib/python2.7/site-packages/yaml/constructor.py", line 415, in construct_undefined
    node.start_mark)
yaml.constructor.ConstructorError: could not determine a constructor for the tag 'tag:yaml.org,2002:python/object/new:detectron.utils.collections.AttrDict'
  in "<unicode string>", line 1, column 1:
    !!python/object/new:detectron.ut ... 

Anyone else have this issue? Thanks.

Hello there, I am interested in developing and adapting DensePose (https://github.com/facebookresearch/detectron2/tree/main/projects/DensePose) into a mobile product. To attract the investment required to 1) adapt DensePose into the required form 2)Port everything onto the mobile 3)Get the design and marketing right, there needs to be at least some commercial upside for the investors.

I am a little confused because detectron2, seems to allow for commercial use https://github.com/facebookresearch/detectron2/blob/main/LICENSE and DensePose is a subproject thereof.

But the original DensePose licence https://github.com/facebookresearch/DensePose/blob/main/LICENSE is Attribution-NonCommercial 4.0 International and quite clearly states that one may "produce, reproduce, and Share Adapted Material for NonCommercial purposes only."

So what does it fall under?

Hello ,

I was wondering if there is any line of work that addresses the problem of generating synthetic densepose data (UV data, etc) by having an SMPL model with texture

Hello,

I am new to this framework. According to document Detectron is only for GPU. What is the issue if i start running on CPU bases computing (Intel Skylake), OS: Debian GNU/Linux. I understand there will be some latency? Are there any other issues

Found Detectron ops lib: /content/anaconda2/lib/python2.7/site-packages/lib/libcaffe2_detectron_ops_gpu.so [E init_intrinsics_check.cc:43] CPU feature avx is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU. [E init_intrinsics_check.cc:43] CPU feature avx2 is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU. [E init_intrinsics_check.cc:43] CPU feature fma is present on your machine, but the Caffe2 binary is not compiled with it. It means you may not get the full speed of your CPU. usage: infer_simple.py [-h] [--cfg CFG] [--wts WEIGHTS] [--output-dir OUTPUT_DIR] [--image-ext IMAGE_EXT] im_or_folder

infer_simple.py: error: unrecognized arguments: frames/100.jpg frames/101.jpg frames/102.jpg frames/103.jpg frames/104.jpg frames/105.jpg frames/106.jpg frames/107.jpg frames/108.jpg frames/109.jpg frames/10.jpg frames/110.jpg frames/111.jpg frames/112.jpg frames/113.jpg frames/114.jpg frames/115.jpg frames/116.jpg frames/117.jpg frames/118.jpg frames/119.jpg frames/11.jpg frames/120.jpg frames/121.jpg frames/122.jpg frames/123.jpg frames/124.jpg frames/125.jpg frames/126.jpg frames/127.jpg frames/128.jpg frames/129.jpg frames/12.jpg frames/130.jpg frames/131.jpg