With the aggressive scaling of semiconductor devices, the increasing complexity of device structure coupled with tighter metrology error budge has driven up Optical Critical Dimension (OCD) time to solution to a critical point. Machine Learning (ML), thanks to its extremely fast turnaround, has been successfully applied in OCD metrology as an alternative solution to the conventional physical modeling. However, expensive and limited reference data or labeled data set necessary for ML to learn from oftentimes leads to overlearning as well as under-learning, limiting its wide adoption. In this paper, we explored techniques that utilize process information to supplement reference data, and synergizing physical modeling with ML. These techniques have been demonstrated to help overcome this limitation of limited reference data with multiple use cases in challenging OCD metrology for advanced semiconductor nodes.