yolov3-tiny model performs strangely

[xiaozhishang]

[Required] Your device (RaspberryPi3, LaptopPC, or other device name):
LaptopPC-NCS2 [Required] Your device's CPU architecture (armv7l, x86_64, or other architecture name):
x86_64 [Required] Your OS (Raspbian, Ubuntu1604, or other os name):
Ubuntu1804，win10 1809 [Required] Details of the work you did before the problem occurred:

I train a yolov3-tiny model with my own dataset. The actual number of objects is 4, so I set [classes = 4, filters=27] in [yolov3-tiny.cfg].
After the training is completed, the model [yolov3-tiny.weights] is obtained. When [Darknet (.weights) -> Tensorflow (.pb)], the actual 4categories in yolov3-tiny.names are set. In [Tensorflow(.pb)->OpenVINO(.bin)_convertion], in [yolo_v3_tiny_changed.json], set ["classes": 4]. This converts the IR model and detects nothing.
But when I change [4] to [80] (in [yolov3-tiny.cfg], set [classes = 80, filters=255], in [yolov3-tiny.names], set 80 categories, In [yolo_v3_tiny_changed.json], set ["classes": 80]). The IR model thus converted has a good detection effect. [Required] Error message:

Classes are set to 4 training models, no objects can be detected, but the same data, classes set to 80 can detect objects properly

[Required] Overview of problems and questions:

Can someone tell me why this is happening?

[sangminns]

Hi xiaozhishang ~

What is your openvino toolkit version?

I also trained yolo-tiny-v3 model with my own data set. And then I updated my openvino version 2019 R1 in Raspberry PI.

In my case, I converted to xml in 2018 R5 version, but I detected object in Raspberry pi in 2019 R1 version. And than it worked fine.

Thank you.

[xiaozhishang]

Hi xiaozhishang ~

What is your openvino toolkit version?

I also trained yolo-tiny-v3 model with my own data set. And then I updated my openvino version 2019 R1 in Raspberry PI.

In my case, I converted to xml in 2018 R5 version, but I detected object in Raspberry pi in 2019 R1 version. And than it worked fine.

Thank you.

hi ksm357

I will updated my openvino version to 2019,and try it again

Thank you.

[xiaozhishang]

Hi xiaozhishang ~

What is your openvino toolkit version?

I also trained yolo-tiny-v3 model with my own data set. And then I updated my openvino version 2019 R1 in Raspberry PI.

In my case, I converted to xml in 2018 R5 version, but I detected object in Raspberry pi in 2019 R1 version. And than it worked fine.

Thank you.

hi ksm357

I upgraded [openvino] to 2019 and the problem was solved.

Thank you.