xiaolonw/adversarial-frcnn
A-Fast-RCNN re-implements adversarial hard-positive generation for object detection using Fast R-CNN in Caffe.
Not currently ranked — collecting fresh signals.
star history
This project re-implements A-Fast-RCNN (CVPR 2017) in Caffe, using adversarial spatial dropout networks to generate hard positive training examples for object detection. The approach modifies Fast R-CNN to include an adversary network that creates occluded and deformed object appearances, forcing the detector to learn robust features. Results show a 3.2 mAP improvement over baseline Fast R-CNN on VOC 2007 using VGG16.
Frequently asked
- What is xiaolonw/adversarial-frcnn?
- A-Fast-RCNN re-implements adversarial hard-positive generation for object detection using Fast R-CNN in Caffe.
- Is adversarial-frcnn open source?
- Yes — xiaolonw/adversarial-frcnn is an open-source project tracked on heatdrop.
- What language is adversarial-frcnn written in?
- xiaolonw/adversarial-frcnn is primarily written in Python.
- How popular is adversarial-frcnn?
- xiaolonw/adversarial-frcnn has 479 stars on GitHub.
- Where can I find adversarial-frcnn?
- xiaolonw/adversarial-frcnn is on GitHub at https://github.com/xiaolonw/adversarial-frcnn.