Sgie(classifier) not attaching the metadata in classifier_meta_list

• Hardware Platform (Jetson / GPU) - GPU
• DeepStream Version - 6.0
• JetPack Version (valid for Jetson only)
• TensorRT Version - 8.0.1
• NVIDIA GPU Driver Version (valid for GPU only) - 515.65.01
• Issue Type( questions, new requirements, bugs) - questions
• 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)
deepstream_test2 is the reference app

I’m unable to access the output of my sgie which is a classifier trained on TAO toolkit. Attaching the pipeline snippet, classifier and sgie config for reference.

The purpose of the classifier is to say whether the detected face from pgie is a real or a spoof image. I tried to access the output of the sgie from the probe function attached at the sinkpad of nvconv just before nvosd.
But obj_meta.classifier_meta_list is always None though pgie gives detection output.
FYI - I have one more sgie in the pipeline whose gie_id is different.
pipeline.txt (6.4 KB)

Please let me know if there is any issue with the pipeline linking or with sgie config.

model_int8.etlt_b1_gpu0_int8.engine (2.8 MB)
liveness_sgie_config.txt (2.2 KB)

please refer to deepstream_python_apps/apps/deepstream-test2 at master · NVIDIA-AI-IOT/deepstream_python_apps · GitHub , first model is detection, the second is classification.

Hey, Sorry for the late update. The issue is resolved when I change the format in the labels.txt file. Whoever encounter this issue in future, please keep the format of the labels.txt as below:

Somehow, semicolon was missed when I generate the engine file from TAO toolkit.

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