Normalization of each output before activation

wojciechmo / yolo2

Train YOLOv2 object detector from scratch using Tensorflow.

138 stars 43 forks source link

Normalization of each output before activation #10

Open DaisyXten opened 6 years ago

DaisyXten commented 6 years ago

From the paper and also other keras implementation of YOLO2, I saw normalization of each layer output. Have you implement this layer or just I missed it somewhere, could you point it out. Thank you a lot

weijinkang commented 5 years ago

Batch normalization is in the function of conv_layer.