details about the resnet 50 structure

Hi, can someone say what does it exactly means. Resnet 50 is 50 layers deep. this 177 is the total number of layers present in network?

I am not able to unnderstan this completely.

Answer 1

[Akshat]

Hi,

As per my understanding of the question, you want to know why are there 177 layers in Resnet 50, where Resnet 50 is a 50 layers deep model.

The reason is that in the 50 layers which are counted in Resnet 50, the pooling layers (max and average) and the activation functions (ReLU) aren’t counted towards the total layers. In order to know what all layers are there in the MATLAB implementation of Resnet50, you can type the following lines in the command window of MATLAB.

net = resnet50();

net.Layers

This is the documentation of Resnet50 in the Deep learning toolbox of MATLAB:

https://www.mathworks.com/help/deeplearning/ref/resnet50.html

You can refer this link to get to know about the layers in general:

https://iq.opengenus.org/resnet50-architecture/

Hope this helps!