(prerequisite-YOLO) One Stage Object Detection

• One Stage Object Detection(Single-Stage Object Detector)

 

Single-Stage ( top ) and Two-Stage Detector ( bottom )

 

A class of object detection architectures that are one-stage. They treat object detection as a simple regression problem; for example, the input image is fed to the network, directly outputs the class probabilities and bounding box coordinates.

These models skip the region proposal stage, also known as Region Proposal Network, which is generally part of Two-Stage Object Detectors that are areas of the image that could contain an object.

 

In single-stage, we apply the detection head directly on the feature map, while, in two-stage, we first apply a region-proposal network on the feature maps.

---

2-Stage Object Detector	1-stage Object Detector
Regional Proposal과 Classification 을 순차적으로 진행하여 객체 검출	영역 추출에 대한 좌표와 이미지 피처를 CNN을 통해 한번에 학습하여 결과를 도출
- 1-stage에 비해 높은 검출 성능	- 2-stage에 비해 낮은 검출 성능
- 1-stage에 비해 속도가 느림	- 2-stage에 비해 속도가 빠름
→ 정확성은 높으나 속도로 인해 실시간 객체탐지가 어려움	→ 실시간 객체 탐지 가능
대표 모델 : R-CNN / Fast R-CNN / Faster R-CNN	대표 모델 : RetinaNet, SSD, EfficientDet, YOLO

 

https://velog.io/@qtly_u/Object-Detection-Architecture-1-or-2-stage-detector-%EC%B0%A8%EC%9D%B4