Statements (23)
| Predicate | Object |
|---|---|
| gptkbp:instanceOf |
computer vision technique
|
| gptkbp:category |
convolutional neural networks
deep learning |
| gptkbp:enables |
end-to-end training of object detectors
|
| gptkbp:fullName |
Region of Interest pooling
|
| https://www.w3.org/2000/01/rdf-schema#label |
RoI pooling
|
| gptkbp:input |
region proposals
feature map |
| gptkbp:introduced |
gptkb:Shaoqing_Ren
gptkb:Ross_Girshick gptkb:Jian_Sun gptkb:Kaiming_He |
| gptkbp:introducedIn |
2015
Faster R-CNN paper |
| gptkbp:output |
fixed-size feature map
|
| gptkbp:purpose |
extract fixed-size feature maps from variable-size regions
|
| gptkbp:relatedTo |
RoI Align
spatial pyramid pooling |
| gptkbp:usedIn |
gptkb:Faster_R-CNN
gptkb:Mask_R-CNN object detection |
| gptkbp:bfsParent |
gptkb:Faster_R-CNN
|
| gptkbp:bfsLayer |
6
|