Issue in running sample TAO application on Jetson orin using DeepStream SDK

[aniketrohara]

I am trying to run the facial landmark sample application on my Jetson Orin using DeepStream SDK and it gets stuck at the following point

[Morganh]

Looks like it is a topic for deepstream.
Moving to deepstream forum.

[yingliu]

Please provide complete information as applicable to your setup.
• Hardware Platform (Jetson / GPU)
• DeepStream Version
• JetPack Version (valid for Jetson only)
• TensorRT Version
• NVIDIA GPU Driver Version (valid for GPU only)
• Issue Type( questions, new requirements, bugs)
• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)
• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)
• The pipeline being used

[aniketrohara]

• Hardware: Jetson AGX Orin
• Deesptream7.0
• Jetpack 6.0 (6.0+b106 and 6.0+b87 both are installed) L4T 36.3.0
• TensorRT 8.6.2
• NVRM version: NVIDIA UNIX Open Kernel Module for aarch64 540.3.0
• Issue : not able to run the facial landmark TAO sample app

[yuweiw]

Could you add the GST_DEBUG=3 in the front of your command and attach the log?

[aniketrohara]

log.txt (1.2 KB)

i redirected the ouput in a txt file

[yuweiw]

Could you attach the ross.jpg to us so that we can try on our side?

[aniketrohara]

sure here it is

[yuweiw]

Our hardware decoder nvjpegdec does not support decoding the image currently. We’ll check that issue. For now you need to create other decoder like jpegdec according to our demo sources\apps\sample_apps\deepstream-test1 yourself.

[aniketrohara]

so wii it work on mp4 input video?

[yuweiw]

Yes. Our nvjpeg decoder just doesn’t support the codec format of your picture. As long as you use ffmpeg to convert your image format, we can also support that.

1. install ffmpeg on your board
2. run the command below:
ffmpeg -i .\ross.jpg dest.jpg
3. use the dest.jpg to run our demo 

[aniketrohara]

Thanks for advice the model worked on a sample input video but as you can see that the output is not appropriate so what are the possible reasons and what can i do now?

a

[yuweiw]

This could be another issue about the inference, not the hardware decoder issue. About this new issue, could you file a new topic? We try to make sure that one topic only solves the same issue, so that others can refer to it easily. Thanks

[aniketrohara]

Okay thank you very much.