To help me understand, are you saying a video lecture is missing ? What makes you think that ?
The lecturer claimed in the end of Video " Simple Convolutional Network Example" (followed by “Pooling layers” on the left side) that he will introduce pooling and fully connected neuron networks. After pooling, he directly jumps to “CNN Example”, using fully connected layers as ‘FC3/4’ in the screenshot.
Prof Ng does explain in the lectures how we add FC layers as the last few layers in a Convolutional Net that is doing classification. But FC layers were already covered in Course 1 of this series, so he doesn’t feel that it is necessary to say that much about them. We already have studied them in detail in Courses 1 and 2. They are the same here as they were there. All you have to do is “flatten” the output of the last Conv or Pooling layer and those vectors become the input to the fully connected network at the end.
Now I see it. Thanks!