Skip to content

Region Proposal Network (RPN) of Faster R-CNN implementation in Python and Tensorflow 1.8 with explanations.

Notifications You must be signed in to change notification settings

Ugenteraan/Faster-RCNN-RPN

Repository files navigation

Region Proposal Network

Region Proposal Network (RPN) is a part of Faster R-CNN algorithm that is used to propose regions (bounding boxes) that contains the targetted objects. These proposed regions are then refined and the objects inside the regions will be classified. RPN can also be used as a standalone network as a one-class object detector. In this implementation, a VGG-16 network was first used to classify between aeroplanes and bicycles. RPN then uses this pre-trained model as the backbone to enable the localization of aeroplanes and bicycles in images.

Below are some of the results of the RPN on localizing airplanes and bicycles.

Region Proposals on Aeroplane images

Region Proposals on Bicycle images

License


MIT

About

Region Proposal Network (RPN) of Faster R-CNN implementation in Python and Tensorflow 1.8 with explanations.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published