[R] Instance-segmentation combined with regression per instance

Hi there. For a project I'm looking into adding a regression head to a Mask R-CNN model. The aim is to predict a continuous variable for each of the detected instances. An example could be to predict the distance to the camera for each of the detected pedestrians. Or the age, weight or height.

I wonder if there is any research into this. And what the name of the problem statement or related field of research is.

I've already did …

instance-segmentation machinelearning regression segmentation