Question 1 of 15easy
Why is VGGNet considered a "deep" network?
Question 2 of 15easy
What is the main purpose of convolution filters in VGGNet?
Question 3 of 15easy
Why does VGGNet mainly use small convolution filters?
Question 4 of 15easy
What is the main role of Max Pooling in VGGNet?
Question 5 of 15easy
In the first layers of VGGNet, the network usually learns:
Question 6 of 15medium
Why are several small convolutions often better than one large convolution?
Question 7 of 15medium
What happens to learned features as we move deeper into VGGNet?
Question 8 of 15medium
What does it mean when we say the receptive field increases in deeper layers?
Question 9 of 15medium
Why is ReLU important in VGGNet?
Question 10 of 15medium
Compared to AlexNet, what was VGGNet mainly trying to improve?
Question 11 of 15hard
Why can deeper VGG layers recognize complex objects while early layers cannot?
Question 12 of 15hard
Why does VGGNet require high computational resources?
Question 13 of 15hard
Why are pre-trained VGG models useful for transfer learning?
Question 14 of 15hard
What problem appears when networks become very deep, leading to architectures like ResNet?
Question 15 of 15hard
Which statement best describes the learning strategy of VGGNet?