22.02.2024 Views

Daniel Voigt Godoy - Deep Learning with PyTorch Step-by-Step A Beginner’s Guide-leanpub

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

• reduce the number of output channels for the max pooling branch.

The 3x3 and 5x5 convolution branches may still output many channels (one for

each filter), but each filter is convolving a reduced number of input channels.

You can think of the RGB-to-grayscale conversion: Instead of using three-channel

convolutions (as in Chapter 6) for colored images, it would use a single-channel

filter (as in Chapter 5) for grayscale (dimension-reduced) images. For a 3x3 filter /

kernel, it means using nine parameters instead of twenty-seven. We can definitely

go deeper!

Inception Modules | 529

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!