FCOS is an anchor-box free, proposal free, single-stage object detection model. By eliminating the predefined set of anchor boxes, FCOS avoids computation related to anchor boxes such as calculating overlapping during training. It also avoids all hyper-parameters related to anchor boxes, which are often very sensitive to the final detection performance.
Source: FCOS: Fully Convolutional One-Stage Object DetectionPaper | Code | Results | Date | Stars |
---|
Task | Papers | Share |
---|---|---|
Object Detection | 54 | 36.73% |
Semantic Segmentation | 13 | 8.84% |
Instance Segmentation | 12 | 8.16% |
Pedestrian Detection | 8 | 5.44% |
Pseudo Label | 4 | 2.72% |
Autonomous Driving | 4 | 2.72% |
Domain Adaptation | 3 | 2.04% |
Pose Estimation | 3 | 2.04% |
Classification | 3 | 2.04% |