googlenet

GoogLeNet

In this directory, we aim to implement the GoogLeNet convolutional neural network (CNN) model for image classification, to be tested with the ImageNet dataset.

In the diagrams subdirectory, we recreated the GoogLeNet architecture diagram in SVG, as well as an annotated version of the architecture with the Inception modules highlighted, and a simplified version with each Inception module collapsed to a single node, which makes it easier to visualize and implement the network. See the directory's documentation for details.

Available implementations:

• Keras

Our implementation is based on the following paper:

• Christian Szegedy, Wei Liu, Yangqing Jia, Pierre Sermanet, Scott E. Reed, Dragomir Anguelov, D. Erhan, Vincent Vanhoucke, and Andrew Rabinovich. “Going deeper with convolutions.” 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) (2015): 1-9.

You can access this paper via:

• arXiv
• Google Scholar

See also:

• Papers with Code