Semantic Segmentation Models
SegFormer
Introduced by Xie et al. in SegFormer: Simple and Efficient Design for Semantic Segmentation with TransformersSegFormer is a Transformer-based framework for semantic segmentation that unifies Transformers with lightweight multilayer perceptron (MLP) decoders. SegFormer has two appealing features: 1) SegFormer comprises a novel hierarchically structured Transformer encoder which outputs multiscale features. It does not need positional encoding, thereby avoiding the interpolation of positional codes which leads to decreased performance when the testing resolution differs from training. 2) SegFormer avoids complex decoders. The proposed MLP decoder aggregates information from different layers, and thus combining both local attention and global attention to render powerful representations.
Source: SegFormer: Simple and Efficient Design for Semantic Segmentation with Transformers
Papers
| Paper | Code | Results | Date | Stars |
|---|
Tasks
| Task | Papers | Share |
|---|---|---|
| Semantic Segmentation | 19 | 36.54% |
| Image Segmentation | 3 | 5.77% |
| Instance Segmentation | 3 | 5.77% |
| Autonomous Driving | 2 | 3.85% |
| Super-Resolution | 2 | 3.85% |
| Change Detection | 2 | 3.85% |
| Image Classification | 2 | 3.85% |
| Object Detection | 2 | 3.85% |
| Lesion Segmentation | 1 | 1.92% |
Mix-FFN
