I was able to convert the yolo model and cfg files to IR through the steps that are documented in the newest release of OpenVino. But i couldn't find a sample that will help me run the converted IR files. I have tried the sample end2end_video_analytics_ie sample but is giving a error. Could you please help me out of the situation.
Link Copied
What errors are you getting? Also are your running on CPU or GPU device? If CPU try the GPU device ( -d GPU ).
I modified the object_detection_demo_ssd_async sample and I can run tiny YOLO v2 without any errors on the GPU device but had errors on the CPU device.
Nikos wrote:
What errors are you getting? Also are your running on CPU or GPU device? If CPU try the GPU device ( -d GPU ).
I modified the object_detection_demo_ssd_async sample and I can run tiny YOLO v2 without any errors on the GPU device but had errors on the CPU device.
Could you please share the source code as mentioned in the above comment
Nikos wrote:
What errors are you getting? Also are your running on CPU or GPU device? If CPU try the GPU device ( -d GPU ).
I modified the object_detection_demo_ssd_async sample and I can run tiny YOLO v2 without any errors on the GPU device but had errors on the CPU device.
Would you mind sharing the modifications you made to run YOLOv2? Thanks
Dear Amalina,
look at this thread for getting more info about Yolo and Inference Engine samples: https://software.intel.com/en-us/forums/computer-vision/topic/785534
OpenVINO does not have a sample adapted for running Yolo, you need to modify one in order to support it.
Best,
Severine
For more complete information about compiler optimizations, see our Optimization Notice.