Vision Models

Explore Vision Models concepts and their connections.

2 concepts · 0 branches

Vision Models

Explore Vision Models concepts and their connections.

2 concepts · 1 sections

Vision Models

CNN
Anchor
Convolutional Neural Network · 卷积神经网络
CNN 用共享卷积核逐层提取局部模式，并把它们组合成更高层视觉特征。
卷积局部模式视觉特征
ViT
Anchor
Vision Transformer · 视觉 Transformer
ViT 把图像切成 Patch Token，再用 Transformer 建模它们之间的全局关系。
图像分类patch视觉模型