Created by: wojtuss

I have a MobileNet-SSD model with trained parameters from Baidu. Also, I have generated a model with the DetectionMAP layer saved. However, I can see the DetectionMAP requires some parameters which are not present with the model from Baidu. These parameters are used for calculation of the accumulative MAP. How the accumulative MAP is calculated? Can it be calculated without those missing parameters? (e.g. as an average of bach MAPs) Do you have also trained parameters for the DetectionMAP layer?