Weakly Supervised Instance Segmentation using Class Peak Response
Yanzhao Zhou, Yi Zhu, Qixiang Ye, Qiang Qiu, and Jianbin Jiao
PRM extracts fine-detailed instance-aware visual cues from class peak responses
PRM enables classification networks for instance mask extraction
PRM is naturally compatible with modern architectures (VGG/ResNet etc.)
PRM introduces negligible training overhead
PRM extracts fine-detailed instance-aware visual cues from class peak responses
Source code and Demo
Paper (arXiv)
Supplementary materials
Poster
Presentation
Thank you for your interest in our work!
Our paper has been accepted by
CVPR 2018
(Spotlight).
The code and demo will be release as soon as possible. You can get instant email notification once we update by
subscribing the repository
.
Got it