I am successfully training on my dataset, but was wondering why is there None as the shape when the model architecture is being listed:
For each row it starts with ‘None’ shouldn’t it be the batch size.
input_image (InputLayer) (None, 3, 480, 840) 0
conv1 (Conv2D) (None, 64, 240, 420) 9408 input_image[0][0]
bn_conv1 (BatchNormalization) (None, 64, 240, 420) 256 conv1[0][0]
activation_1 (Activation) (None, 64, 240, 420) 0 bn_conv1[0][0]
block_1a_conv_1 (Conv2D) (None, 64, 120, 210) 36864 activation_1[0][0]
block_1a_bn_1 (BatchNormalizati (None, 64, 120, 210) 256 block_1a_conv_1[0][0]
activation_2 (Activation) (None, 64, 120, 210) 0 block_1a_bn_1[0][0]
block_1a_conv_2 (Conv2D) (None, 64, 120, 210) 36864 activation_2[0][0]
block_1a_conv_shortcut (Conv2D) (None, 64, 120, 210) 4096 activation_1[0][0]
block_1a_bn_2 (BatchNormalizati (None, 64, 120, 210) 256 block_1a_conv_2[0][0]
block_1a_bn_shortcut (BatchNorm (None, 64, 120, 210) 256 block_1a_conv_shortcut[0][0]
add_1 (Add) (None, 64, 120, 210) 0 block_1a_bn_2[0][0]
block_1a_bn_shortcut[0][0]
activation_3 (Activation) (None, 64, 120, 210) 0 add_1[0][0]
block_1b_conv_1 (Conv2D) (None, 64, 120, 210) 36864 activation_3[0][0]
block_1b_bn_1 (BatchNormalizati (None, 64, 120, 210) 256 block_1b_conv_1[0][0]
activation_4 (Activation) (None, 64, 120, 210) 0 block_1b_bn_1[0][0]
block_1b_conv_2 (Conv2D) (None, 64, 120, 210) 36864 activation_4[0][0]
block_1b_conv_shortcut (Conv2D) (None, 64, 120, 210) 4096 activation_3[0][0]
block_1b_bn_2 (BatchNormalizati (None, 64, 120, 210) 256 block_1b_conv_2[0][0]
block_1b_bn_shortcut (BatchNorm (None, 64, 120, 210) 256 block_1b_conv_shortcut[0][0]
add_2 (Add) (None, 64, 120, 210) 0 block_1b_bn_2[0][0]
block_1b_bn_shortcut[0][0]
activation_5 (Activation) (None, 64, 120, 210) 0 add_2[0][0]
block_1c_conv_1 (Conv2D) (None, 64, 120, 210) 36864 activation_5[0][0]
block_1c_bn_1 (BatchNormalizati (None, 64, 120, 210) 256 block_1c_conv_1[0][0]
activation_6 (Activation) (None, 64, 120, 210) 0 block_1c_bn_1[0][0]
block_1c_conv_2 (Conv2D) (None, 64, 120, 210) 36864 activation_6[0][0]
block_1c_conv_shortcut (Conv2D) (None, 64, 120, 210) 4096 activation_5[0][0]
block_1c_bn_2 (BatchNormalizati (None, 64, 120, 210) 256 block_1c_conv_2[0][0]
block_1c_bn_shortcut (BatchNorm (None, 64, 120, 210) 256 block_1c_conv_shortcut[0][0]
add_3 (Add) (None, 64, 120, 210) 0 block_1c_bn_2[0][0]
block_1c_bn_shortcut[0][0]
activation_7 (Activation) (None, 64, 120, 210) 0 add_3[0][0]
block_2a_conv_1 (Conv2D) (None, 128, 60, 105) 73728 activation_7[0][0]