fail to run faster-rcnn

Hi
I am trying to run faster-rcnn using deepstream-app -c deepstream_app_config_fasterRCNN.txt,
I am getting this error: ERROR from primary_gie_classifier: Failed to create NvDsInferContext instance
any idea how to figure it out?

Hi,

Have you followed the instructions mentioned in:
/opt/nvidia/deepstream/deepstream-4.0/sources/objectDetector_FasterRCNN/README

You will need to build the custom library before launching TensorRT.
If you do follow the steps, would you mind to share the complete log with us so we can check it further?

Thanks.

I am new to deepstream. I follow up the process, but I do not have custom library.
where to start?

Hi,

The custom library can be compiled with this command:

$ make -C nvdsinfer_custom_impl_fasterRCNN

You will also need to download the fasterRCNN model with following commands:

$ wget --no-check-certificate https://dl.dropboxusercontent.com/s/o6ii098bu51d139/faster_rcnn_models.tgz?dl=0 -O faster-rcnn.tgz
$ tar zxvf faster-rcnn.tgz -C . --strip-components=1 --exclude=ZF_*
$ cp /usr/src/tensorrt/data/faster-rcnn/faster_rcnn_test_iplugin.prototxt .

For more details, please check our README file:
/opt/nvidia/deepstream/deepstream-4.0/sources/objectDetector_FasterRCNN/README

Thanks.

Thanks Aasta. it works.
I want to implement CRNN on deepstream. is there any documentation or guideline?

Hi,

You can check this document for the customized model implementation:
https://docs.nvidia.com/metropolis/deepstream/dev-guide/index.html#page/DeepStream_Development_Guide%2Fdeepstream_custom_model.html%23wwpID0ERHA

Thanks.