Convolutional Neural Networks
Atrous-convolution block
Introduced by Zhou et al. in ACNN: a Full Resolution DCNN for Medical Image SegmentationAtrous Convolution Neural Network (ACNN), as a pooling-free network structure, is proposed to achieve full-resolution feature processing using a theoretically optimal dilation setting for a larger receptive field, even with fewer parameters. Compared to other techniques, it can achieve higher segmentation Intersection over Union (IoU) and much less trainable parameters and model sizes, indicating the benefit of full-resolution feature maps in feature processing.
Source: ACNN: a Full Resolution DCNN for Medical Image Segmentation
Papers
| Paper | Code | Results | Date | Stars |
|---|
Tasks
| Task | Papers | Share |
|---|---|---|
| Antibody-antigen binding prediction | 1 | 14.29% |
| Deep Learning | 1 | 14.29% |
| Protein Function Prediction | 1 | 14.29% |
| Computed Tomography (CT) | 1 | 14.29% |
| Image Segmentation | 1 | 14.29% |
| Medical Image Segmentation | 1 | 14.29% |
| Semantic Segmentation | 1 | 14.29% |
Usage Over Time
Components
| Component | Type |
|
|---|---|---|
| 🤖 No Components Found | You can add them if they exist; e.g. Mask R-CNN uses RoIAlign |
