Lesson 2 of 4

CNNs for vision

Watch, read the card, then practise. CNN architectures · Transfer learning

Video coming soon

The video for this lesson is being recorded.

Meanwhile, the notes below and the 60-second cards cover the same ideas.

CNN architectures
Vision CNNs stack convolution and pooling blocks: the image shrinks while the number of features grows.
Remember: Deeper isn't always better: pick a model that fits your data and speed needs.
Transfer learning
Transfer learning starts from a model trained on millions of images and adapts it to your task.
Remember: It's the default for small image datasets.

Try the example in Code Lab

size, channels = 224, 64
for block in range(3):
    size, channels = size // 2, channels * 2
print(size, channels)

Lesson 2 of 4

Watch, read the card, then practise. CNN architectures · Transfer learning

Video coming soon

The video for this lesson is being recorded.

Meanwhile, the notes below and the 60-second cards cover the same ideas.

CNN architectures
Vision CNNs stack convolution and pooling blocks: the image shrinks while the number of features grows.
Remember: Deeper isn't always better: pick a model that fits your data and speed needs.
Transfer learning
Transfer learning starts from a model trained on millions of images and adapts it to your task.
Remember: It's the default for small image datasets.

size, channels = 224, 64
for block in range(3):
    size, channels = size // 2, channels * 2
print(size, channels)