anyone managed to make use of movidius VPU to accelerate the auto annotation of CVAT? hardly can find information in Internet for this.
I found that the CVAT docker can be build with openvino support but so far only support to work on CPU (seems like the opencv CVAT team doesn't have a movidius stick/VPU to enable the support)
Hi Kim Chuan,
Greetings to you.
CVAT auto annotation uses only CPU plugin.
And here is the CVAT auto annotation inference_engine.py link for your reference.