When i run deepstream-test5, it shows wrong like this:

i use deepstream-5.1 on jetson NX, and kafka server is ok because i can run test4 successfully.
i work on rtsp stream from IP camera,
i note in this file: dstest5_msgconv_sample_config, sensor type=Camera, should i change it to type=3 which same as PGIE?
pls give some advices to run test-5.

thank you very much!

2021-07-06 05-37-25屏幕截图

You do not need to modify this file, these are the predefined values. can you share the configuration file you used?

test5_wang.txt (7.8 KB)

this is my txt.file about test5.
it shows erro like this:
sgs-001@sgs001-desktop:~/opt/nvidia/deepstream/deepstream-5.1/sources/apps/sample_apps/deepstream-test5-wang$ ./deepstream-test5-app -c test5_wang.txt -p 0
Failed to load config file: No such file or directory
** ERROR: <gst_nvinfer_parse_config_file:1260>: failed

Using winsys: x11
%3|1625545322.917|FAIL|rdkafka#producer-1| [thrd:test-server:9092/0]: test-server:9092/0: Failed to resolve ‘test-server:9092’: Name or service not known
%3|1625545322.917|ERROR|rdkafka#producer-1| [thrd:test-server:9092/0]: test-server:9092/0: Failed to resolve ‘test-server:9092’: Name or service not known
WARNING: [TRT]: Using an engine plan file across different models of devices is not recommended and is likely to affect performance or even cause errors.
0:00:03.803766042 12274 0x55a08bb530 INFO nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<secondary_gie_2> NvDsInferContext[UID 6]: Info from NvDsInferContextImpl::deserializeEngineAndBackend() <nvdsinfer_context_impl.cpp:1702> [UID = 6]: deserialized trt engine from :/home/sgs-001/opt/nvidia/deepstream/deepstream-5.1/sources/apps/sample_apps/deepstream-test5-wang/Secondary_Car/Secondary_CarMake/resnet18.caffemodel_b16_gpu0_int8.engine
INFO: [Implicit Engine Info]: layers num: 2
0 INPUT kFLOAT input_1 3x224x224
1 OUTPUT kFLOAT predictions/Softmax 20x1x1

0:00:03.804020837 12274 0x55a08bb530 INFO nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger:<secondary_gie_2> NvDsInferContext[UID 6]: Info from NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:1806> [UID = 6]: Use deserialized engine model: /home/sgs-001/opt/nvidia/deepstream/deepstream-5.1/sources/apps/sample_apps/deepstream-test5-wang/Secondary_Car/Secondary_CarMake/resnet18.caffemodel_b16_gpu0_int8.engine
0:00:03.804148986 12274 0x55a08bb530 ERROR nvinfer gstnvinfer.cpp:613:gst_nvinfer_logger:<secondary_gie_2> NvDsInferContext[UID 6]: Error in NvDsInferContextImpl::preparePreprocess() <nvdsinfer_context_impl.cpp:917> [UID = 6]: Cannot access mean image file ‘/home/sgs-001/opt/nvidia/deepstream/deepstream-5.1/sources/apps/sample_apps/deepstream-test5-wang/Secondary_Car/Secondary_CarMake/…/…/models/Secondary_CarMake/mean.ppm’
ERROR: Infer Context prepare preprocessing resource failed., nvinfer error:NVDSINFER_CONFIG_FAILED
0:00:03.809067968 12274 0x55a08bb530 WARN nvinfer gstnvinfer.cpp:812:gst_nvinfer_start:<secondary_gie_2> error: Failed to create NvDsInferContext instance
0:00:03.809186902 12274 0x55a08bb530 WARN nvinfer gstnvinfer.cpp:812:gst_nvinfer_start:<secondary_gie_2> error: Config file path: /home/sgs-001/opt/nvidia/deepstream/deepstream-5.1/sources/apps/sample_apps/deepstream-test5-wang/Secondary_Car/Secondary_CarMake/config_infer_secondary_carmake.txt, NvDsInfer Error: NVDSINFER_CONFIG_FAILED
** ERROR: main:1451: Failed to set pipeline to PAUSED
Quitting
ERROR from secondary_gie_2: Failed to create NvDsInferContext instance
Debug info: /dvs/git/dirty/git-master_linux/deepstream/sdk/src/gst-plugins/gst-nvinfer/gstnvinfer.cpp(812): gst_nvinfer_start (): /GstPipeline:pipeline/GstBin:secondary_gie_bin/GstNvInfer:secondary_gie_2:
Config file path: /home/sgs-001/opt/nvidia/deepstream/deepstream-5.1/sources/apps/sample_apps/deepstream-test5-wang/Secondary_Car/Secondary_CarMake/config_infer_secondary_carmake.txt, NvDsInfer Error: NVDSINFER_CONFIG_FAILED

Please make sure infer configuration file path correct.
like this, /home/sgs-001/opt/nvidia/deepstream/deepstream-5.1/sources/apps/sample_apps/deepstream-test5-wang/Secondary_Car/Secondary_CarMake/config_infer_secondary_carmake.txt
others similar.

sir,
i runs original path by command:
./deepstream-test5-app -c configs/XXXXX(original long name file),
test5 works successfull already!

but i should build my own project for test5 because original paths are too crazy and troubles,
i put all related files to my own path name is: deepstream_test5_wang, and all files used absolute path…but it faild still.
i think it needs some tools which i have to put them in my path, what do they should be?
now i am ensure that’s path problem only…
pls give some advices, thank you very much!

If all the file path you use is absolute path, no reason it fail. please check if any other failure.

wait i am checking…

thank you very much, sir
it is path problem and no need to use absolute path at all.
i got my own test5 project to run successfully already,
and i change to my custom models to get it work.

thanks!

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.