Going deeper with convolutions

Christian Szegedy(Google (United States)), Wei Liu(University of North Carolina at Chapel Hill), Yangqing Jia(IGlobal University), Pierre Sermanet(IGlobal University), Scott Reed(University of Michigan–Ann Arbor), Dragomir Anguelov(Google (United States)), Dumitru Erhan(Google (United States)), Vincent Vanhoucke(IGlobal University), Andrew Rabinovich(Magic Leap (United States))
Unknown
June 1, 2015
10.1109/cvpr.2015.7298594
Cited by 46,713

Abstract

We propose a deep convolutional neural network architecture codenamed Inception that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14). The main hallmark of this architecture is the improved utilization of the computing resources inside the network. By a carefully crafted design, we increased the depth and width of the network while keeping the computational budget constant. To optimize quality, the architectural decisions were based on the Hebbian principle and the intuition of multi-scale processing. One particular incarnation used in our submission for ILSVRC14 is called GoogLeNet, a 22 layers deep network, the quality of which is assessed in the context of classification and detection.

Related Papers

ImageNet classification with deep convolutional neural networks
Alex Krizhevsky, Ilya Sutskever, Geoffrey E. Hinton|Communications of the ACM|2017|75.7k
Gradient-based learning applied to document recognition
Yann LeCun, Léon Bottou, Yoshua Bengio et al.|Proceedings of the IEEE|1998|57.9k
Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
Ross Girshick, Jeff Donahue, Trevor Darrell et al.|Unknown|2014|31.6k
Visualizing and Understanding Convolutional Networks
Matthew D. Zeiler, Rob Fergus|Lecture notes in computer science|2014|15.4k
Backpropagation Applied to Handwritten Zip Code Recognition
Yann LeCun, Bernhard E. Boser, J. S. Denker et al.|Neural Computation|1989|11.8k