When getting classifier meta in deepstream-test2, it is always null

Please provide complete information as applicable to your setup.

**• Hardware Platform (Jetson / GPU) Jetson
**• DeepStream Version 5.0
**• JetPack Version (valid for Jetson only) 4.4 DP[L4T 32.4.2]
**• TensorRT Version 7.1.0.16
**• NVIDIA GPU Driver Version (valid for GPU only) CUDA 10.2.89
**• 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)
I run example deepstream-test2 in deepstream_python_apps/apps/deepstream-test2 at master · NVIDIA-AI-IOT/deepstream_python_apps · GitHub
I want to get classifier_meta, so I edit deepstream_test_2.py like following, but I found that I could not get classifier_meta_list, it’s always None.
Is there anything wrong?

Hey, can you disable the classifier-async-mode=1 and try again?

I have set classifier-async-mode=0 in dstest2_sgie1_config.txt/dstest2_sgie2_config.txt/dstest2_sgie3_config.txt, but I still could not get classifier_meta_list, it’s still always None

I use the following files to test:
deepstream_test_2.py (12.2 KB)
dstest2_pgie_config.txt (3.4 KB)
dstest2_sgie1_config.txt (3.5 KB) dstest2_sgie2_config.txt (3.5 KB) dstest2_sgie3_config.txt (3.5 KB)
dstest2_tracker_config.txt (1.7 KB)

I just add the following code following your code in original deepsteram python test2 and it’s not always null, which model/stream are you using?

                # Casting l_obj.data to pyds.NvDsObjectMeta
                obj_meta=pyds.NvDsObjectMeta.cast(l_obj.data)
                if obj_meta.classifier_meta_list is None:
                    print("none none none none none none none none none none none none none none none ")
                else:
                    print("ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ")
            except StopIteration:


Frame Number=296 Number of Objects=5 Vehicle_count=2 Person_count=2
none none none none none none none none none none none none none none none
ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok
none none none none none none none none none none none none none none none
none none none none none none none none none none none none none none none
ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok
none none none none none none none none none none none none none none none
none none none none none none none none none none none none none none none
none none none none none none none none none none none none none none none
none none none none none none none none none none none none none none none
Frame Number=297 Number of Objects=9 Vehicle_count=5 Person_count=3
none none none none none none none none none none none none none none none
none none none none none none none none none none none none none none none
none none none none none none none none none none none none none none none
ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok
ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok
ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok
none none none none none none none none none none none none none none none
none none none none none none none none none none none none none none none
none none none none none none none none none none none none none none none
none none none none none none none none none none none none none none none
none none none none none none none none none none none none none none none
Frame Number=298 Number of Objects=11 Vehicle_count=7 Person_count=3
^Cnone none none none none none none none none none none none none none none
none none none none none none none none none none none none none none none
ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok
ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok
ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok  ok
none none none none none none none none none none none none none none none
none none none none none none none none none none none none none none none
none none none none none none none none none none none none none none none
none none none none none none none none none none none none none none none
none none none none none none none none none none none none none none none
none none none none none none none none none none none none none none none
Frame Number=299 Number of Objects=11 Vehicle_count=6 Person_count=4

I used the model configured in dstest2_pgie_config.txt/dstest2_sgie1_config.txt/dstest2_sgie2_config.txt/dstest2_sgie3_config.txt.
e.g. model in dstest2_pgie_config.txt

The stream I used is
sample_720p.h264 (14.1 MB)

Yeah, same as me.
Could you only apply the chang on my last comment on pure deepstream python test2 and try if you can get the same result as mine?

Yes, I applied your last comment on pure deepstream test2, but met the smae issue.
I think my pyds.so may be not correct. I did not find it in my Jetson before, so got it from a deepstream image(I forgot which deepstream image, sorry)

Could you help give me your pyds.so(Jetson deepstream 5.0)? I will try it

Hey customer,
I think it’s better you can use the latest DS5.1 and follow Quickstart Guide — DeepStream 5.1 Release documentation and GitHub - NVIDIA-AI-IOT/deepstream_python_apps: A project demonstrating use of Python for DeepStream sample apps given as a part of SDK (that are currently in C,C++).

Follow README instructions in the release package to install. Note that the bindings are now installed as part of the SDK. Please see the HOW-TO Guide Guide for details.

Thanks, I will try DS5.1 instead of DS5.0

I have tried DS5.1, it works now. thanks a lot.
So this issue seems to be a bug of DS5.0, and DS5.1 fixs it

Hey customer, we don’t find this bug on DS5.0 in our internal SQA test, so I guess maybe it related to setup or something else.

Get it, thanks