Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Xception: Deep Learning with Depthwise Separable Convolutions #7

Open
guanfuchen opened this issue Nov 22, 2018 · 3 comments
Open

Xception: Deep Learning with Depthwise Separable Convolutions #7

guanfuchen opened this issue Nov 22, 2018 · 3 comments

Comments

@guanfuchen
Copy link
Owner

related paper

摘要
We present an interpretation of Inception modules in convolutional neural networks as being an intermediate step in-between regular convolution and the depthwise separable convolution operation (a depthwise convolution followed by a pointwise convolution). In this light, a depthwise separable convolution can be understood as an Inception module with a maximally large number of towers. This observation leads us to propose a novel deep convolutional neural network architecture inspired by Inception, where Inception modules have been replaced with depthwise separable convolutions. We show that this architecture, dubbed Xception, slightly outperforms Inception V3 on the ImageNet dataset (which Inception V3 was designed for), and significantly outperforms Inception V3 on a larger image classification dataset comprising 350 million images and 17,000 classes. Since the Xception architecture has the same number of parameters as Inception V3, the performance gains are not due to increased capacity but rather to a more efficient use of model parameters.
@guanfuchen
Copy link
Owner Author

overview about the idea

image

image

@guanfuchen
Copy link
Owner Author

result

classification on ImageNet, size, speed with Inception V3

image

residual importance

image

@guanfuchen
Copy link
Owner Author

conclusion

image

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant