Error while Running Pose_Estimation in DeepStream container

[skrjha20]

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU) → GPU
• DeepStream Version → 6.2
**• NVIDIA GPU Driver Version (valid for GPU only) → 525.85.12, GPU: Nvidia Tesla T4

We are using DeepStream Container and trying to run pose_estimation application using

Creating a Human Pose Estimation Application with NVIDIA DeepStream | NVIDIA...

We are inside the container. We have git cloned the pose_estimation code and places it under sample_apps folder in DeepStream container

root@56952d815a18:/opt/nvidia/deepstream/deepstream-6.2/sources/apps/sample_apps/deepstream_pose_estimation#

We have edited the Makefile and changed the NVDS_VERSION:=5.0 to NVDS_VERSION:=6.2
Apart from this there is no other changes done

Now we want to compile this app using the command given in Step 6
$ sudo make
and we are getting the error
bash: sudo: command not found

Kindly help

[fanzh]

in docker, please only use “make”.

[skrjha20]

Thanks for the command

We have successfully complied the using make command.

It has generated the deepstream-pose-estimation-app file shown below

Next we run the deepstream_pose_estimation_app inside container using the command
./deepstream-pose-estimation-app ./images/input.gif ./images/
We are using the same input file which was available in the git

NVIDIA-AI-IOT/deepstream_pose_estimation/blob/master/images/input.gif

When it ran 1st time, it took around 3-4 mins and gave the error But we got the error

But it generated the output file in the image folder Pose_Estimation.mp4

I thought may be the input file is not h264 type, so I downloaded the h264 type file from http://docs.evostream.com/sample_content/assets/bun33s.mp4

and placed at images folder

I ran the command
./deepstream-pose-estimation-app ./images/bun33s.mp4 ./images/
but got the same error

Please help to solve this error.
Can you help me with the exact input file I should use

[fanzh]

please try ./deepstream-pose-estimation-app /opt/nvidia/deepstream/deepstream/samples/streams/sample_720p.h264 ./images/

[skrjha20]

Thanks for the command

We have successfully run the deepstream_pose_estimation_app and it has generated the output in images as Pose_Estimation.mp4 file
But we are not able to play this output file in container. Also we are not able to play the sample_720p.h264 file in container

Can you please let us know how to play these files in container.
Also since pose_estimation needs .h264 files to process, can you please guide us which camera we should use to generate streaming videos of .h264 vidoes.

Also I believe the camera will have an IP address which we can use for pose estimation for streaming videos.
Kindly help with queries

[fanzh]

There is no update from you for a period, assuming this is not an issue any more.
Hence we are closing this topic. If need further support, please open a new one.
Thanks

@skrjha20

But we are not able to play this output file in container. Also we are not able to play the sample_720p.h264 file in container

do you mean you can’t play that Pose_Estimation.mp4? can you play by windows’s VLC player? sample_720p.h264’s content is the same with /opt/nvidia/deepstream/deepstream/samples/streams/sample_720p.mp4.

@skrjha20

Also I believe the camera will have an IP address which we can use for pose estimation for streaming videos.

deepstream_pose_estimation is only a sample, if want to use mp4 or RTSP, you can modify the code, please refer to \opt\nvidia\deepstream\deepstream-6.2\sources\apps\sample_apps\deepstream-test3\deepstream_test3_app.c, it use nvurisrcbin or uridecodebin to play MP4 or RTSP.