Table 1. A brief history of CNNs in Image Segmentation.

2014: R-CNN	An Early Application of CNNs to ObjectDetection : propose a bunch of boxes in the image and see if any of them actually correspond to an object
2015: Fast R-CNN	Speeding up and Simplifying R-CNNInsight 1: RoI (Region of Interest) PoolingInsight 2: Combine All Models into One Network
2016: Faster R-CNN	Speeding Up Region Proposal : only one CNN needs to be trained
2017: Mask R-CNN	Extending Faster R-CNN for Pixel Level Segmentation : RoiAlign - Realigning RoIPool to be More Accurate