A Brief History of CNNs in Image Segmentation: From R-CNN to Mask R-CNN
Dhruv Parthasarathy
1K17

Thank you for this awesome post! I have a question regarding the implementation of Faster-RCNN. You said there is no need for Region Proposals as the input, but the implementations I’ve seen (e.g. https://github.com/yhenon/keras-frcnn) need annotations as inputs as well. I thought the Region-Proposing Network does this in the network for us?