$\color{black}\rule{365px}{3px}$

Link to Paper:

“Rethinking the Inception Architecture for Computer Vision” - 2015

arxiv.org

Table of Contents

1. Introduction

$\color{black}\rule{365px}{3px}$

Christian Szegedy | Vincent Vanhoucke | *Sergey Ioffe *| Jonathon Shlens

Contributions

• Developed idea of GoogleNet(Inception-V1)
• What is so special about GoogleNet?
  • Although VGGNet has the compelling feature of architectural simplicity, this comes at a high cost: evaluating the network requires a lot of computation.
  • On the other hand, the Inception architecture of GoogLeNet was also designed to perform well even under strict constraints on memory and computational budget.
• Improvements from Inception-V1:
  • Use of BatchNormalization
  • Asymmetric Factorized Convolutions