Learning-Based Local Visual Representation and Indexing

Chapter 1

Introduction

Abstract

With the rapid advances in multimedia, communication, and Internet technology, today there is an ever-increasing amount of multimedia content with rich metadata tags on the Web. Of this massive multimedia data, image and video data represent major proportions and contain rich content information, which is very accessible for human observation and manipulation. Image and video datasets are also widely used in medical analysis, digital libraries, computer vision, and intelligent vehicles. Consequently, how to organize such digital media toward effective indexing and search is recognized as one of the fundamental challenges in the above fields, with a bright future in both academic and industrial research. Consequently, ...