Issue with integrating deepstream faciallandmarks app Deepstream 6.2 dGPU

Deepstream 6.2 on dGPU

Im trying to integrate the deepstream landmarks app into my other application, by that I mean, Im trying to take the required code in deepstream faciallandmarks app and putting in my application. It get compiled but while running im getting a seg fault, the logs are

Thread 10 "trois-frs" received signal SIGSEGV, Segmentation fault.
[Switching to Thread 0x7fff3109f000 (LWP 548425)]
0x00007ffff3a14e21 in cvcore::faciallandmarks::FacialLandmarksPostProcessor::execute(cvcore::Array<cvcore::ArrayN<cvcore::Vector2<float>, 200ul> >&, cvcore::Tensor<(cvcore::TensorLayout)1, (cvcore::ChannelCount)4, (cvcore::ChannelType)5> const&, cvcore::Array<cvcore::BBox> const&, CUstream_st*) () from /opt/nvidia/deepstream/deepstream-6.2/lib/cvcore_libs/libnvcv_faciallandmarks.so
(gdb) bt
#0  0x00007ffff3a14e21 in cvcore::faciallandmarks::FacialLandmarksPostProcessor::execute(cvcore::Array<cvcore::ArrayN<cvcore::Vector2<float>, 200ul> >&, cvcore::Tensor<(cvcore::TensorLayout)1, (cvcore::ChannelCount)4, (cvcore::ChannelType)5> const&, cvcore::Array<cvcore::BBox> const&, CUstream_st*) () at /opt/nvidia/deepstream/deepstream-6.2/lib/cvcore_libs/libnvcv_faciallandmarks.so
#1  0x0000555555596c4e in landmark_buffer_probe(_GstPad*, _GstPadProbeInfo*, void*) (pad=0x55555682d670 [GstPad|src], info=0x7fff3106a4e0, u_data=0x0) at src/prediction_probe.cpp:1626
#2  0x00007ffff389945e in  () at /lib/x86_64-linux-gnu/libgstreamer-1.0.so.0
#3  0x00007ffff36d7996 in g_hook_list_marshal () at /lib/x86_64-linux-gnu/libglib-2.0.so.0
#4  0x00007ffff3898ade in  () at /lib/x86_64-linux-gnu/libgstreamer-1.0.so.0
#5  0x00007ffff389cf1a in  () at /lib/x86_64-linux-gnu/libgstreamer-1.0.so.0
#6  0x00007ffff38a3e63 in gst_pad_push () at /lib/x86_64-linux-gnu/libgstreamer-1.0.so.0
#7  0x00007fff6133bf39 in gst_nvinfer_output_loop(void*) () at /usr/lib/x86_64-linux-gnu/gstreamer-1.0/deepstream/libnvdsgst_infer.so
#8  0x00007ffff3711ae1 in  () at /lib/x86_64-linux-gnu/libglib-2.0.so.0
#9  0x00007fffed223609 in start_thread (arg=<optimized out>) at pthread_create.c:477
#10 0x00007ffff285d133 in clone () at ../sysdeps/unix/sysv/linux/x86_64/clone.S:95

any idea on what can cause this ?

also, the deepstream faciallandmark app works fine when I run it standalone.

My application runs if I do not attach the probe function landmark_buffer_probe from deepstream-faciallandmark-app.

I solved it. Thanks anyways.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.