**• Hardware Platform Jetson Nano
**• DeepStream Version: Deepstream 5.0
**• JetPack Version (valid for Jetson only): 4.4
I trained a model detectnet_v2_resnet18.etlt with my own dataset by tlt-train in this container nvcr.io/nvidia/tlt-streamanalytics:v2.0_py3, it show the accuracy is 90%, but when i put the etlt model to deepstream python apps, nothing can be detected, I have tried with deepstream-test1 and deepstream-imagedata-multistream-redaction, it is running but no bounding box comes out…
I have adjust the configure files there already, not sure what the problems are, thanks!
Please have a try with latest deepstream 6.0. Thanks.
The thing is when I try with Deepstream 6.0, there are always a lot plugins etc. shows missing. And deepstream 5.0 should do the work as well, right? Deepstream 5.0 works now with test-1 and imagedata-multistream-redaction, but not nvdsanalytics, the error is as below it seems no tracker lib can be found:
root@everisai-nano:/opt/nvidia/deepstream/deepstream-5.0/sources/python/apps/deepstream-nvdsanalytics-copy# GST_DEBUG=3 python3 deepstream_nvdsanalytics.py file:///opt/nvidia/deepstream/deepstream/sources/python/supermarket_KB.h264
Creating Pipeline
Creating streamux
Creating source_bin 0
Creating source bin
source-bin-00
Creating Pgie
Creating nvtracker
Creating nvdsanalytics
Creating tiler
Creating nvvidconv
Creating nvosd
Creating transform
Creating EGLSink
Adding elements to Pipeline
Linking elements in the Pipeline
Now playing…
1 : file:///opt/nvidia/deepstream/deepstream/sources/python/supermarket_KB.h264
Starting pipeline
Using winsys: x11
gstnvtracker: Loading low-level lib at /opt/nvidia/deepstream/deepstream/lib/libnvds_nvmultiobjecttracker.so
gstnvtracker: Failed to open low-level lib at /opt/nvidia/deepstream/deepstream/lib/libnvds_nvmultiobjecttracker.so
dlopen error: /opt/nvidia/deepstream/deepstream/lib/libnvds_nvmultiobjecttracker.so: cannot open shared object file: No such file or directory
gstnvtracker: Failed to initilaize low level lib.
0:00:01.380635891 2835 0x1ee14070 ERROR nvtracker gstnvtracker.cpp:238:gst_nv_tracker_start: gstnvtracker: Failed to initialize trackerIface
0:00:01.380690735 2835 0x1ee14070 WARN GST_PADS gstpad.c:1149:gst_pad_set_active:tracker:sink Failed to activate pad
gstnvtracker: Loading low-level lib at /opt/nvidia/deepstream/deepstream/lib/libnvds_nvmultiobjecttracker.so
gstnvtracker: Failed to open low-level lib at /opt/nvidia/deepstream/deepstream/lib/libnvds_nvmultiobjecttracker.so
dlopen error: /opt/nvidia/deepstream/deepstream/lib/libnvds_nvmultiobjecttracker.so: cannot open shared object file: No such file or directory
gstnvtracker: Failed to initilaize low level lib.
0:00:01.386501674 2835 0x1e307cf0 ERROR nvtracker gstnvtracker.cpp:238:gst_nv_tracker_start: gstnvtracker: Failed to initialize trackerIface
0:00:01.386560789 2835 0x1e307cf0 WARN GST_PADS gstpad.c:1149:gst_pad_set_active:tracker:sink Failed to activate pad
gstnvtracker: Loading low-level lib at /opt/nvidia/deepstream/deepstream/lib/libnvds_nvmultiobjecttracker.so
gstnvtracker: Failed to open low-level lib at /opt/nvidia/deepstream/deepstream/lib/libnvds_nvmultiobjecttracker.so
dlopen error: /opt/nvidia/deepstream/deepstream/lib/libnvds_nvmultiobjecttracker.so: cannot open shared object file: No such file or directory
gstnvtracker: Failed to initilaize low level lib.
0:00:01.387166101 2835 0x1e307cf0 ERROR nvtracker gstnvtracker.cpp:238:gst_nv_tracker_start: gstnvtracker: Failed to initialize trackerIface
0:00:01.387258705 2835 0x1e307cf0 WARN GST_PADS gstpad.c:1149:gst_pad_set_active:tracker:sink Failed to activate pad
0:00:01.387293237 2835 0x1e307cf0 WARN bin gstbin.c:2832:reset_state: Failed to switch back down to PAUSED
0:00:13.825735088 2835 0x1e307cf0 INFO nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger: NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::deserializeEngineAndBackend() <nvdsinfer_context_impl.cpp:1701> [UID = 1]: deserialized trt engine from :/opt/nvidia/deepstream/deepstream-5.0/samples/models/tlt_pretrained_models/knife_model_Dec3/resnet18_detector_Dec3.etlt_b1_gpu0_fp16.engine
INFO: [Implicit Engine Info]: layers num: 3
0 INPUT kFLOAT input_1 3x1280x720
1 OUTPUT kFLOAT output_bbox/BiasAdd 8x80x45
2 OUTPUT kFLOAT output_cov/Sigmoid 2x80x45
ERROR: [TRT]: INVALID_ARGUMENT: Cannot find binding of given name: conv2d_bbox
0:00:13.870356206 2835 0x1e307cf0 WARN nvinfer gstnvinfer.cpp:616:gst_nvinfer_logger: NvDsInferContext[UID 1]: Warning from NvDsInferContextImpl::checkBackendParams() <nvdsinfer_context_impl.cpp:1669> [UID = 1]: Could not find output layer ‘conv2d_bbox’ in engine
ERROR: [TRT]: INVALID_ARGUMENT: Cannot find binding of given name: conv2d_cov/Sigmoid
0:00:13.870467977 2835 0x1e307cf0 WARN nvinfer gstnvinfer.cpp:616:gst_nvinfer_logger: NvDsInferContext[UID 1]: Warning from NvDsInferContextImpl::checkBackendParams() <nvdsinfer_context_impl.cpp:1669> [UID = 1]: Could not find output layer ‘conv2d_cov/Sigmoid’ in engine
0:00:13.870595945 2835 0x1e307cf0 INFO nvinfer gstnvinfer.cpp:619:gst_nvinfer_logger: NvDsInferContext[UID 1]: Info from NvDsInferContextImpl::generateBackendContext() <nvdsinfer_context_impl.cpp:1805> [UID = 1]: Use deserialized engine model: /opt/nvidia/deepstream/deepstream-5.0/samples/models/tlt_pretrained_models/knife_model_Dec3/resnet18_detector_Dec3.etlt_b1_gpu0_fp16.engine
0:00:14.672322848 2835 0x1e307cf0 INFO nvinfer gstnvinfer_impl.cpp:313:notifyLoadModelStatus: [UID 1]: Load new model:dsnvanalytics_pgie_config.txt sucessfully
0:00:14.888080676 2835 0x1e307cf0 WARN basesrc gstbasesrc.c:3583:gst_base_src_start_complete:
Decodebin child added: source
Decodebin child added: decodebin0
0:00:14.993064590 2835 0x1e307cf0 WARN basesrc gstbasesrc.c:3583:gst_base_src_start_complete:
Decodebin child added: h264parse0
Decodebin child added: capsfilter0
Decodebin child added: nvv4l2decoder0
Opening in BLOCKING MODE
0:00:18.807902530 2835 0x1e307cf0 WARN v4l2 gstv4l2object.c:4430:gst_v4l2_object_probe_caps:nvv4l2decoder0:src Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:00:18.808020082 2835 0x1e307cf0 WARN v4l2 gstv4l2object.c:2372:gst_v4l2_object_add_interlace_mode:0x7ef9874010 Failed to determine interlace mode
NvMMLiteOpen : Block : BlockType = 261
NVMEDIA: Reading vendor.tegra.display-size : status: 6
NvMMLiteBlockCreate : Block : BlockType = 261
0:00:18.924177812 2835 0x1e307cf0 WARN v4l2 gstv4l2object.c:4430:gst_v4l2_object_probe_caps:nvv4l2decoder0:src Failed to probe pixel aspect ratio with VIDIOC_CROPCAP: Unknown error -1
0:00:18.924268750 2835 0x1e307cf0 WARN v4l2 gstv4l2object.c:2372:gst_v4l2_object_add_interlace_mode:0x7ef9874010 Failed to determine interlace mode
In cb_newpad
gstname= video/x-raw
features= <Gst.CapsFeatures object at 0x7f81fe01c8 (GstCapsFeatures at 0x7ef9838a40)>
0:00:18.987312645 2835 0x1e307cf0 WARN v4l2videodec gstv4l2videodec.c:1632:gst_v4l2_video_dec_decide_allocation: Duration invalid, not setting latency
0:00:19.004222656 2835 0x1e307cf0 WARN v4l2bufferpool gstv4l2bufferpool.c:1057:gst_v4l2_buffer_pool_start:nvv4l2decoder0:pool:src Uncertain or not enough buffers, enabling copy threshold
0:00:19.026621942 2835 0x7ef9ddf370 WARN v4l2bufferpool gstv4l2bufferpool.c:1503:gst_v4l2_buffer_pool_dqbuf:nvv4l2decoder0:pool:src Driver should never set v4l2_buffer.field to ANY
but for Deepstream 6.0, nothing works yet… as below:
Deepstream 6.0 has the error as below even with deepstream-test-1:
root@everisai-nano:/opt/nvidia/deepstream/deepstream-6.0/sources/python/apps/deepstream-test1# GST_DEBUG=3 python3 deepstream_test_1.py /opt/nvidia/deepstream/deepstream-5.0/samples/streams/sample_720p.h264
Traceback (most recent call last):
File “deepstream_test_1.py”, line 28, in
import pyds
ModuleNotFoundError: No module named ‘pyds’
that’s why I’d like to go forwards with 5.0, since it started to work with some apps, but only how to solve the issue in nvdsanalytics… thanks a lot!
it started to work now with deepstream 5.0, thanks for the help.
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.