Position-Sensitive RoI Pooling layer aggregates the outputs of the last convolutional layer and generates scores for each RoI. Unlike RoI Pooling, PS RoI Pooling conducts selective pooling, and each of the $k$ × $k$ bin aggregates responses from only one score map out of the bank of $k$ × $k$ score maps. With end-to-end training, this RoI layer shepherds the last convolutional layer to learn specialized position-sensitive score maps.
Source: R-FCN: Object Detection via Region-based Fully Convolutional NetworksPaper | Code | Results | Date | Stars |
---|
Task | Papers | Share |
---|---|---|
Object Detection | 24 | 36.92% |
General Classification | 3 | 4.62% |
Object Recognition | 2 | 3.08% |
Traffic Sign Detection | 2 | 3.08% |
Semantic Segmentation | 2 | 3.08% |
Real-Time Object Detection | 2 | 3.08% |
Image Classification | 2 | 3.08% |
Classification | 2 | 3.08% |
Object Tracking | 1 | 1.54% |
Component | Type |
|
---|---|---|
🤖 No Components Found | You can add them if they exist; e.g. Mask R-CNN uses RoIAlign |