Training a customized object detection algorithm for opemMV

Hello.
I am trying to train a customised object detection algorithm for OpenMV.
I am wondering is there any tutorial or link to help me to do so?
Is it even possible?
Any help is highly appreciated