Some Improvements on Deep Convolutional Neural Network Based Image Classification

19 Dec 2013Andrew G. Howard

We investigate multiple techniques to improve upon the current state of the art deep convolutional neural network based image classification pipeline. The techiques include adding more image transformations to training data, adding more transformations to generate additional predictions at test time and using complementary models applied to higher resolution images... (read more)

PDF Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK LEADERBOARD
Image Classification ImageNet Five Base + Five HiRes Top 1 Accuracy 66.3% # 145
Top 5 Accuracy 86.3% # 105

Methods used in the Paper