Full Description
This ground-breaking text/reference diverges
from the traditional view that computer vision (for image analysis) and string
processing (for text mining) are separate and unrelated fields of study,
propounding that images and text can be treated in a similar manner for the
purposes of information retrieval, extraction and classification. Highlighting
the benefits of knowledge transfer between the two disciplines, the text
presents a range of novel similarity-based learning (SBL) techniques founded on
this approach. Topics and features: describes a variety of SBL approaches,
including nearest neighbor models, local learning, kernel methods, and
clustering algorithms; presents a nearest neighbor model based on a novel
dissimilarity for images; discusses a novel kernel for (visual) word
histograms, as well as several kernels based on a pyramid representation; introduces
an approach based on string kernels for native language identification; contains
links for downloading relevant open source code.
Contents
Motivation and Overview.- Learning Based on Similarity.- Part I: Knowledge Transfer from Text Mining to Computer Vision.- State of the Art Approaches for Image Classification.- Local Displacement Estimation of Image Patches and Textons.- Object Recognition with the Bag of Visual Words Model.- Part II: Knowledge Transfer from Computer Vision to Text Mining.- State of the Art Approaches for String and Text Analysis.- Local Rank Distance.- Native Language Identification with String Kernels.- Spatial Information in Text Categorization.- Conclusions.
