Information Fusion | springerprofessional.de

Springer Professional

nach oben

2022 | Buch

Kapitel lesen Erstes Kapitel lesen

Information Fusion

Machine Learning Methods

verfasst von: Dr. Jinxing Li, Prof. Bob Zhang, Prof. David Zhang

Verlag: Springer Nature Singapore

Enthalten in: Springer Professional "Wirtschaft+Technik" , Springer Professional "Technik" , Springer Professional "Wirtschaft"

Inhaltsverzeichnis

Über dieses Buch

In the big data era, increasing information can be extracted from the same source object or scene. For instance, a person can be verified based on their fingerprint, palm print, or iris information, and a given image can be represented by various types of features, including its texture, color, shape, etc. These multiple types of data extracted from a single object are called multi-view, multi-modal or multi-feature data. Many works have demonstrated that the utilization of all available information at multiple abstraction levels (measurements, features, decisions) helps to obtain more complex, reliable and accurate information and to maximize performance in a range of applications.

This book provides an overview of information fusion technologies, state-of-the-art techniques and their applications. It covers a variety of essential information fusion methods based on different techniques, including sparse/collaborative representation, kernel strategy, Bayesian models, metric learning, weight/classifier methods, and deep learning. The typical applications of these proposed fusion approaches are also presented, including image classification, domain adaptation, disease detection, image restoration, etc.

This book will benefit all researchers, professionals and graduate students in the fields of computer vision, pattern recognition, biometrics applications, etc. Furthermore, it offers a valuable resource for interdisciplinary research.

Anzeige

Inhaltsverzeichnis

Frontmatter

Chapter 1. Introduction

Abstract

“Information Fusion” plays a key role in many fields, e.g., machine learning, deep learning, and pattern recognition etc. It is capable of fusing multiple features, modalities, views or algorithms, greatly contributing to the performance improvement in many applications. This chapter introduces what is “Information Fusion”, reviews the history of information fusion and analyzes the main contributions of this book. After reading this chapter, people will have some shallow ideas on information fusion.

Jinxing Li, Bob Zhang, David Zhang

Chapter 2. Information Fusion Based on Sparse/Collaborative Representation

Abstract

Sparse representation follows the insight of data representation of human beings, allowing the data to be more accurately and robustly represented. Recently, many works in image classification, image retrieval, image recovery, etc., have shown its effectiveness. In contrast to the sparse representation, collaborative representation ignores the robustness but encourages algorithms to enjoy a fast computation. This chapter respectively proposes an information fusion method based on the sparse representation and two information fusion methods based on the collaborative representation. After reading this chapter, people can have preliminary knowledge on sparse/collaborative representation based fusion algorithms.

Jinxing Li, Bob Zhang, David Zhang

Chapter 3. Information Fusion Based on Gaussian Process Latent Variable Model

Abstract

Gaussian Process Latent Variable Model (GPLVM) is capable of representing the data without a determined function, which is a generative and non-parametric model. Compared with sparse/collaborative representation, GPLVM enjoys the non-linearity, which does exist in real-world datasets. This chapter proposes three GPLVM based information fusion methods, contributing to the classification performance improvement. After reading this chapter people can have preliminary knowledge on GPLVM based fusion algorithms.

Jinxing Li, Bob Zhang, David Zhang

Chapter 4. Information Fusion Based on Multi-View and Multi-Feature Learning

Abstract

In many real-world applications, the same object can be represented with multiple modalities or views, and each modality can be further represented with various features. This kind of data is often named as multi-view and multi-feature data. To address this data, two probabilistic and generative fusion methods are studied in this chapter. After reading this chapter people, can have preliminary knowledge on Bayesian theory-based fusion algorithms.

Jinxing Li, Bob Zhang, David Zhang

Chapter 5. Information Fusion Based on Metric Learning

Abstract

Metric learning aims to measure the similarity or dissimilarity between each pair of samples. Until now, various types of metrics are studied and achieve satisfied performances in many applications. To comprehensively exploit the advantages of different metrics, this chapter proposes two metric fusion methods and applies them to classification and verification. After reading this chapter people can have preliminary knowledge on metric learning based fusion algorithms.

Jinxing Li, Bob Zhang, David Zhang

Chapter 6. Information Fusion Based on Score/Weight Classifier Fusion

Abstract

By applying different classifiers to the classification, multiple classification scores will be generated. Generally, different classifiers enjoy their own advantages and disadvantages, and only a simple classifier usually fails to robustly achieve the classification. To jointly exploit the advantages of different classifiers, it is significant to use the score fusion. This chapter proposes two score fusion methods and applies them to classification. After reading this chapter, people can have preliminary knowledge on score fusion algorithms.

Jinxing Li, Bob Zhang, David Zhang

Chapter 7. Information Fusion Based on Deep Learning

Abstract

Recently deep learning has achieved outstanding performances in many fields. Compared with conventional shallow models, deep learning methods use deeper architectures to more powerfully model the complex distributions of the real-world datasets. This chapter proposes two deep learning based fusion methods that can fuse two branches of networks into a unique feature. After reading this chapter people can have preliminary knowledge on deep learning based fusion methods.

Jinxing Li, Bob Zhang, David Zhang

Chapter 8. Conclusion

Abstract

In this book, we focus on information fusion and propose several machine learning and deep learning based methods for pattern recognition.

Jinxing Li, Bob Zhang, David Zhang

Backmatter

Titel: Information Fusion
verfasst von: Dr. Jinxing Li
Prof. Bob Zhang
Prof. David Zhang
Copyright-Jahr: 2022
Verlag: Springer Nature Singapore
Electronic ISBN: 978-981-16-8976-5
Print ISBN: 978-981-16-8975-8
DOI: https://doi.org/10.1007/978-981-16-8976-5

Premium Partner