TFlite object detection demo on RB5
Join Date: 8 Jun 21
Posts: 4
Posted: Fri, 2022-01-21 08:59
Hi,
I can't get the Object detection with TensorFlow Lite demo running on the RB5.
I have downloaded everything and it seems to have compiled, however, the program does not run correctly. A window with a video feed is displayed without the object detection. The following error message is shown:
Loading model: mobilenet_quant_v1_224.tflite
(process:17830): GStreamer-CRITICAL **: 16:23:46.387: gst_element_get_state: assertion 'GST_IS_ELEMENT (element)' failed
OpenCV | GStreamer warning: GStreamer: unable to query pipeline state (/home/src/opencv-wayland/modules/videoio/src/cap_gstreamer.cpp:421)
labels finished
ERROR: Subgraph is already in use.
ERROR: Subgraph is already in use.
[*] DEBUG: handle_toplevel_configure: maximized=0 fullscreen=0 resizing=0 focused=0 size=[0 x 0]
ERROR: Invoke called on model that is not ready.
ERROR: Invoke called on model that is not ready.
[*] DEBUG: handle_toplevel_configure: maximized=0 fullscreen=0 resizing=0 focused=1 size=[0 x 0]
ERROR: Invoke called on model that is not ready.
I am using an RB5 board which has been updated using apt-get with a Microsoft webcam.
Thanks in advance for any help.
Dear customer,
Will invoke IOT guys to give mofe feedback to here later.
BR.
Wei
Please check whether you had multiple gstreamer processors: ps -A
Please upload your gstreamer command.
Did you solve this issue? I am having the same problem.