[Deep Learning][CNN][VGG]

[peternara]

• Very Deep Convolutional Networks for Large-Scale Image Recognition (VGG network)
• 특징
  • CNN
  • 224x224 input RGB image
  • very small 3×3 filters in all convolutional layer
  • case vgg16 구조
    • (2 conv layer + 1 max pooling layer)x2+(3 conv layer + 1 max pooling layer)x3 + 3개의 fc layers+softmax layer
  • relu activate function
  • depth to 16–19, weight layers.
    • vgg 16, vgg 19
  • alexnet과 마찬가지로 fc layer에서 deep feature
    • sparse & semantic
  • alexnet보다 약 10%이상의 성능 향상을 보여줌
  • 매우 강건한 network