Deepface 当中提到的Locally connected layer (local convolution)

DeepFace: Closing the Gap to Human-Level Performance in Face Verification

• Face Alignment/Frontalization
  • fiducial point detector
  • 2D alignment
  • 3D alignment
• Feature Extraction
  • conv & maxpooing

    • However, several levels of pooling would cause the network to lose information about the precise position of detailed facial structure and micro-textures. Hence, we apply max-pooling only to the first convolutional layer

  • Locally connected layer

Make full use of aligned INPUT

https://prateekvjoshi.com/2016/04/12/understanding-locally-connected-layers-in-convolutional-neural-networks/