Thanks Katsuya, this is super helpful!
I was able to run this and get ~35 FPS on CPU, ~10 FPS on NCS2 and ~5 FPS on NCS1.
Probably possible to accelerate this further by using more than a single async inference request.
I gave up on tiny-yolov3 +NCS2 until I see your post. Many thanks Katsuya.
I have an application that use tiny-yolov2 with custom data set (4 classes) that needed to speed up the processing time with NCS2. By following your githut. I manage to get it working with CPU. However, was not able to get it working with NCS2 (with MYRIAD).
The steps used are:
Step 1: cover weight.python convert_weights_pb_car.py --weights_fil
e yolov3-tiny_416x416_3800.weights --class_names carcoco.names --size 416 --data_format NHWC
Step 2: convert to TensorFlow model
python convert_weights_pb.py --weights_file yolov3-tiny_416x416_3800.weights --class_names carcoco.names --size 416 --data_format NHWC
Step 3: Run object_detection_demo_yolov3_async
object_detection_demo_yolov3_async -i C:\Users\trafficcountuser\Downloads\OpenVINO-YoloV3-master\data\input\testvideo3.mp4 -m C:\Intel\computer_vision_sdk_2018.5.445\deployment_tools\model_optimizer\frozen_car_yolov3_model.xml -d MYRIAD
As shown at the attached picture, the result is not right for MYRIAD option with random boxes. If used CPU with --data_type=FP32 the result is correct.
I can't figure out the reason why it is work with CPU but not MYRIAD. Any suggestions to making custom data set to work with tiny-YoloV3 + NCS2 would be greatly appreciated.
- Windows 10
- OpenVINO toolkit 2018 R5 (2018.5.445)
- Python 3.6
- tiny-YoloV3 (MS-COCO)
- video file
- custom data set with 4 classes
Katsuya, Thanks for taking the time to helps out.
I added all the files that are needed to reproduce the issue to attached zip file. A readme file is also include to list the commands used and modification done to each files. Please let me if additional clarification are needed.
Could you let me know the quality of detection is reasonable on MYRIAD option when you try it? I see random results on my custom train data set while the result is reasonable on CPU.
Thanks a lot for taking the time to investigate this issue. It's good to know that the issue is limited to my program. I'm hopping someone from Intel with more inside of openVino will have a chance to look at the issue.
Did anybody able to use NCS2 on custom data set on any model ?
You are right. The issue have not been solved. I'm hopping that Intel can have a look at it.
At the mean time, I'm trying to find another model to work with NCS2.