hamidriasat / DDRNets

Deep Dual-resolution Networks for Real-time and Accurate Semantic Segmentation of Road Scenes

MIT License

31 stars 2 forks source link

Operands could not be broadcast together with shapes #2

Closed tricoffee closed 3 years ago

tricoffee commented 3 years ago

Operands could not be broadcast together with shapes (512, 1024, 64) (128, 256, 64)

Low to High

x_temp = layers.Activation("relu")(layers_inside[2]) x_temp = layers.Conv2D(highres_planes, kernel_size=(1,1), use_bias=False)(x_temp) x_temp = layers.BatchNormalization()(x_temp) x_temp = tf.image.resize(x_temp, (height_output, widthoutput)) # 1/16 -> 1/8 x = layers.Add()([x_, x_temp]) # next high branch input, 1/8

tricoffee commented 3 years ago

Are you sure your code can work ???????????????????????????????????????????

hamidriasat commented 3 years ago

Hi tricoffee,

I have confirmed my code is working with input size of [1024, 2048,3] and [512, 1024, 3]. if you are facing problem's kindly share your complete settings so that i can recreate the error on my end. Also make sure you are using TensorFlow 2.0 or later version.

Thanks

hamidriasat commented 3 years ago

Not responding.