Regularization
SpatialDropout
Introduced by Tompson et al. in Efficient Object Localization Using Convolutional NetworksSpatialDropout is a type of dropout for convolutional networks. For a given convolution feature tensor of size $n_{\text{feats}}$×height×width, we perform only $n_{\text{feats}}$ dropout trials and extend the dropout value across the entire feature map. Therefore, adjacent pixels in the dropped-out feature map are either all 0 (dropped-out) or all active as illustrated in the figure to the right.
Source: Efficient Object Localization Using Convolutional Networks
Papers
| Paper | Code | Results | Date | Stars |
|---|
Tasks
| Task | Papers | Share |
|---|---|---|
| Semantic Segmentation | 14 | 21.54% |
| Autonomous Driving | 5 | 7.69% |
| Image Segmentation | 3 | 4.62% |
| Quantization | 2 | 3.08% |
| Autonomous Vehicles | 2 | 3.08% |
| Real-Time Semantic Segmentation | 2 | 3.08% |
| Deep Learning | 2 | 3.08% |
| regression | 2 | 3.08% |
| Medical Image Analysis | 2 | 3.08% |
Usage Over Time
Components
| Component | Type |
|
|---|---|---|
| 🤖 No Components Found | You can add them if they exist; e.g. Mask R-CNN uses RoIAlign |
