GoogLeNet

Introduced by Szegedy et al. in Going Deeper with Convolutions

GoogLeNet is a type of convolutional neural network based on the Inception architecture. It utilises Inception modules, which allow the network to choose between multiple convolutional filter sizes in each block. An Inception network stacks these modules on top of each other, with occasional max-pooling layers with stride 2 to halve the resolution of the grid.

Source: Going Deeper with Convolutions

Read Paper See Code

Papers

Paper	Code	Results	Date	Stars

Tasks

Task	Papers	Share
General Classification	25	12.69%
Image Classification	23	11.68%
Classification	18	9.14%
Quantization	13	6.60%
Object Detection	13	6.60%
Object Recognition	7	3.55%
Domain Adaptation	5	2.54%
Clustering	5	2.54%
Benchmarking	3	1.52%

Usage Over Time

This feature is experimental; we are continuously improving our matching algorithm.

Components

Component	Type	Add Remove
Auxiliary Classifier	Miscellaneous Components
Average Pooling	Pooling Operations
Dense Connections	Feedforward Networks
Dropout	Regularization
Inception Module	Image Model Blocks
Local Response Normalization	Normalization
Max Pooling	Pooling Operations
ReLU	Activation Functions
Softmax	Output Functions

Categories

Add Remove

Convolutional Neural Networks