I am detecting real-time face by using raspberrypi4 model B+, getting 8-10fps with support of the OpenVino toolkit, here I am using a pre-trained IR model. Can anyone please suggest, how can i increase fps?

[RandallMan_B_Intel]

Hello AVerm26,

The frames per second (fps) you are getting depend on multiple variables: the sample demo, model, and resolution that you are using. It's also possible that the delay is being done by displaying the results with OpenCV.

Please, answer the following questions:

• What version of OpenVINO are you using?
• Are you using our face detection demo and model, or yours?
• Are you using the inference engine or OpenCV dnn library?

Regards,

Randall B.