I bought a hailo raspberry pi development-kit and want to try different AI model for people detection.
I know that a lot of pre-trained detection models is evaluated already and I can find the benchmark of mAPs and fps.
My question is that how can I evaluate these models by my own dataset (annotated with vott format)?
Just like calculate mAPs .
It will be so appreciate if there is a python script to evaluate different models.
Hello everyone
I bought a hailo raspberry pi development-kit and want to try different AI model for people detection. I know that a lot of pre-trained detection models is evaluated already and I can find the benchmark of mAPs and fps.
My question is that how can I evaluate these models by my own dataset (annotated with vott format)? Just like calculate mAPs . It will be so appreciate if there is a python script to evaluate different models.
Thank you very much.