docker run --net=host -e DISPLAY --device=/dev/nvidia0    --runtime=nvidia -v /home/jp/Downloads:/data  -it openpose:gpu -e DISPLAY --device=/dev/video0   bash


root@HP-Z440:/openpose# ./build/examples/openpose/openpose.bin -display=0 -video /data/mov.mp4 -face=true -face_net_resolution="368x368" -net_resolution="-1x640" -write_json="/data/kp_368_640" -write_video="/data/face_368_net_640.avi"
Starting OpenPose demo...
Configuring OpenPose...
Starting thread(s)...
Auto-detecting all available GPUs... Detected 1 GPU(s), using 1 of them starting at GPU 0.
F0709 01:00:38.762663    42 syncedmem.cpp:71] Check failed: error == cudaSuccess (2 vs. 0)  out of memory
*** Check failure stack trace: ***
    @     0x7f8d74b565cd  google::LogMessage::Fail()
    @     0x7f8d74b58433  google::LogMessage::SendToLog()
    @     0x7f8d74b5615b  google::LogMessage::Flush()
    @     0x7f8d74b58e1e  google::LogMessageFatal::~LogMessageFatal()
    @     0x7f8d74139c48  caffe::SyncedMemory::mutable_gpu_data()
    @     0x7f8d73fbe052  caffe::Blob<>::mutable_gpu_data()
    @     0x7f8d7419e758  caffe::CuDNNConvolutionLayer<>::Forward_gpu()
    @     0x7f8d74117a31  caffe::Net<>::ForwardFromTo()
    @     0x7f8d761a2be0  op::NetCaffe::forwardPass()
    @     0x7f8d760f7c73  op::FaceExtractorCaffe::forwardPass()
    @     0x7f8d760f34e5  op::WFaceExtractorNet<>::work()
    @     0x7f8d76108ba9  op::Worker<>::checkAndWork()
    @     0x7f8d76108d4b  op::SubThread<>::workTWorkers()
    @     0x7f8d7611181b  op::SubThreadQueueInOut<>::work()
    @     0x7f8d7610bafb  op::Thread<>::threadFunction()
    @     0x7f8d75a3bc80  (unknown)
    @     0x7f8d7518d6ba  start_thread
    @     0x7f8d754aa41d  clone
    @              (nil)  (unknown)