VGG16 and Efficient Models Report

Introduction

Object recognition from images is one of the most basic computer vision problems and has a vast range of uses in almost all fields. Precise identification of objects within the digital images is useful in areas like driverless cars, robotics, security and surveillance systems, image database recognition, etc. The research question formulated for this report was to review and compare various strategies in image object recognition, preferencing the deep learning methodology and testing its efficiency on the Caltech 256 data set.