concept separable convolution in category deep learning

This is an excerpt from Manning's book Deep Learning with R. Login to get full access to this book.

Finally, in 2014, 2015, and 2016, even more advanced ways to help gradient propagation were discovered, such as batch normalization, residual connections, and depthwise separable convolutions. Today we can train from scratch models that are thousands of layers deep.

Figure 7.14. Depthwise separable convolution: a depthwise convolution followed by a pointwise convolution

When it comes to larger-scale models, depthwise separable convolutions are the basis of the Xception architecture, a high-performing convnet that comes packaged with Keras. You can read more about the theoretical grounding for depthwise separable convolutions and Xception in François’s paper “Xception: Deep Learning with Depthwise Separable Convolutions.”^[8]

This is an excerpt from Manning's book Deep Learning with Python. Login to get full access to this book.

Finally, in 2014, 2015, and 2016, even more advanced ways to help gradient propagation were discovered, such as batch normalization, residual connections, and depthwise separable convolutions. Today we can train from scratch models that are thousands of layers deep.

Figure 7.16. Depthwise separable convolution: a depthwise convolution followed by a pointwise convolution

When it comes to larger-scale models, depthwise separable convolutions are the basis of the Xception architecture, a high-performing convnet that comes packaged with Keras. You can read more about the theoretical grounding for depthwise separable convolutions and Xception in my paper “Xception: Deep Learning with Depthwise Separable Convolutions.”^[8]