2019-05-28-VGG（Very Deep Convolu

• investigate the effect of the convolutional network depth on its
  accuracy

• a thorough evaluation of networks of increasing depth using an architecture with very small (3×3) convolution filters

算法框架

• 3个3×3卷积核相比1个7×7卷积核的优势
  • make the decision function morediscriminative by incorporating three non-linear rectification layers
  • decrease the number of parameters
• 网络D比C更具优势的原因
  • while the additional non-linearity does help (C is better than B), it is also important to capture spatial context by using conv. filters with non-trivial receptive fields (D is better than C)