Content-Aware ReAssembly of FEatures (CARAFE) is an operator for feature upsampling in convolutional neural networks. CARAFE has several appealing properties: (1) Large field of view. Unlike previous works (e.g. bilinear interpolation) that only exploit subpixel neighborhood, CARAFE can aggregate contextual information within a large receptive field. (2) Content-aware handling. Instead of using a fixed kernel for all samples (e.g. deconvolution), CARAFE enables instance-specific content-aware handling, which generates adaptive kernels on-the-fly. (3) Lightweight and fast to compute.
Source: CARAFE: Content-Aware ReAssembly of FEaturesPaper | Code | Results | Date | Stars |
---|
Task | Papers | Share |
---|---|---|
Object Detection | 3 | 27.27% |
Instance Segmentation | 2 | 18.18% |
Semantic Segmentation | 2 | 18.18% |
Depth Estimation | 1 | 9.09% |
Monocular Depth Estimation | 1 | 9.09% |
Panoptic Segmentation | 1 | 9.09% |
Feature Upsampling | 1 | 9.09% |
Component | Type |
|
---|---|---|
🤖 No Components Found | You can add them if they exist; e.g. Mask R-CNN uses RoIAlign |