Mobilenetv2
서론
[작성중…]
영상 분할 , 물체와 경계를 찾는데 사용된다.
1.
use lightweight depthwise convolutions to filter features as
a source of non-linearity
2.
important to remove non-linearities in the narrow layers in order to maintain
representational power
3.
입력과 출력의 형태가 같은 decoupling 구조
4.
Classification : ImageNet Classification
Object Detection : COCO Object Detection
Image Segmentation : VOC Image Segmentation
5.
manifold learning
https://roytravel.tistory.com/105