Coverart for item
The Resource Advances in Chinese document and text processing, edited by Cheng-Lin Liu, Yue Lu

Advances in Chinese document and text processing, edited by Cheng-Lin Liu, Yue Lu

Label
Advances in Chinese document and text processing
Title
Advances in Chinese document and text processing
Statement of responsibility
edited by Cheng-Lin Liu, Yue Lu
Contributor
Subject
Language
eng
Summary
"The book is a collection of invited chapters by experts in Chinese document and text processing, and is part of a series on Language Processing, Pattern Recognition, and Intelligent Systems. The chapters introduce the latest advances and state-of-the-art methods for Chinese document image analysis and recognition, font design, text analysis and speaker recognition. Handwritten Chinese character recognition and text line recognition are at the core of document image analysis (DIA), and therefore, are addressed in four chapters for different scripts (online characters, offline characters, ancient characters, and text lines). Two chapters on character recognition pay much attention to deep convolutional neural networks (CNNs), which are widely used and performing superiorly in various pattern recognition problems. A chapter is contributed to describe a large handwriting database consisting both online and offline characters and text pages. Postal mail reading and writer identification, addressed in two chapters, are important applications of DIA. The collection can serve as reference for students and engineers in Chinese document and text processing and their applications."--Publisher's website
Member of
Dewey number
005
Illustrations
illustrations
Index
index present
Language note
In English, with examples in Chinese
Literary form
non fiction
Nature of contents
  • dictionaries
  • bibliography
http://library.link/vocab/relatedWorkOrContributorName
  • Liu, Cheng-Lin,
  • Lu, Yue,
Series statement
Series on language processing, pattern recognition, and intelligent systems
Series volume
vol. 2
http://library.link/vocab/subjectName
  • Text processing (Computer science)
  • Chinese language
Label
Advances in Chinese document and text processing, edited by Cheng-Lin Liu, Yue Lu
Instantiates
Publication
Copyright
Bibliography note
Includes bibliographical references and index
Contents
  • Preface; Chapter 1 Characteristics of English, Chinese, and Arabic Typefaces; 1. Introduction; 1.1. The purpose of this chapter; 1.2. Literature review of typeface personality studies; 1.3. Participants; 1.4. Materials; 1.5. Results; 2. Chinese Character Characteristics Compared with English; 2.1. Overview; 2.2. Relative baseline feature; 2.3. Squared character feature; 2.4. Weight detection; 2.5. Serif / sans serif / script; 2.6. Structure design; 2.7. Stroke contrast; 2.8. Design proportion; 2.9. Width to height ratio; 2.10. Character spacing; 2.11. Counter design & inner design
  • 2.12. Stem and cap height design2.13. Junction points; 2.14. Stroke ending; 2.15. Curve; 3. Comparing Arabic with Latin Font Characteristics; 3.1. Overview; 3.2. Feature of Arabic script compared to Latin script; 3.3. Anatomy of characters; 3.4. Design considerations; References; Chapter 2 Chinese Handwriting Database Building and Benchmarking; 1. Introduction; 2. Data Collection Settings; 2.1. Character sets; 2.2. Data collection; 3. Data Annotation; 3.1. Annotation of offline data; 3.2. Annotation of online data; 3.3. Data format; 4. Statistics of Datasets; 5. Recommendations of Usage
  • 5.1. Data partitioning5.2. Research scenarios; 5.2.1. Handwritten document segmentation; 5.2.2. Handwritten character recognition; 5.2.3. Text line recognition; 5.2.4. Handwritten document retrieval; 5.2.5. Writer adaptation; 5.2.6. Writer identi cation; 6. Preliminary Evaluation; 7. Competition Results; 8. Conclusion; References; Chapter 3 CNN Based Handwritten Character Recognition; 1. Introduction; 1.1. Development of deep learning; 1.2. CNN for image understanding; 1.3. Character recognition by CNN; 2. Overview of the CNN-Based Handwritten Character Recognition System
  • 2.1. The framework of the system2.2. Recognition beyond human; 3. Random Distortion for Sample Generation; 4. Training Tricks of CNN Model; 5. Model Scale and Input Image Size; 6. Multi-Model Voting; 7. Conclusion; References; Chapter 4 Online Handwritten Chinese Character Recognition: From a Bayesian Approach to Deep Learning; 1. Introduction; 2. Online HCCR under a Bayesian Statistical Learning Framework; 2.1. General framework for OHCCR; 2.2. Preprocessing; 2.2.1. Normalization; 2.2.2. Data augmentation with deformation transformation; 2.2.3. Imaginary stroke technique
  • 2.3. Feature extraction2.3.1. Region partition with the meshing technique; 2.3.2. 8-directional feature extraction; 2.3.3. Signature of path features; 2.4. Feature dimension reduction with LDA; 2.5. MQDF classifier; 3. OHCCR-CNN: An End-to-End Approach for OHCCR using Deep Convolutional Neural Networks; 3.1. Brief introduction to CNNs; 3.1.1. Convolutional layer; 3.1.2. Pooling layer; 3.1.3. Softmax layer and loss function; 3.1.4. Platforms for training CNNs; 3.2. Domain knowledge-enhanced DCNN for OHCCR; 3.3. Efficient training of CNN using a new algorithm named DropSample
Control code
ocn981975306
Dimensions
unknown
Extent
1 online resource
Form of item
online
Isbn
9789813143685
Note
World Scientific Publishing
Other physical details
illustrations
Specific material designation
remote
System control number
(OCoLC)981975306
Label
Advances in Chinese document and text processing, edited by Cheng-Lin Liu, Yue Lu
Publication
Copyright
Bibliography note
Includes bibliographical references and index
Contents
  • Preface; Chapter 1 Characteristics of English, Chinese, and Arabic Typefaces; 1. Introduction; 1.1. The purpose of this chapter; 1.2. Literature review of typeface personality studies; 1.3. Participants; 1.4. Materials; 1.5. Results; 2. Chinese Character Characteristics Compared with English; 2.1. Overview; 2.2. Relative baseline feature; 2.3. Squared character feature; 2.4. Weight detection; 2.5. Serif / sans serif / script; 2.6. Structure design; 2.7. Stroke contrast; 2.8. Design proportion; 2.9. Width to height ratio; 2.10. Character spacing; 2.11. Counter design & inner design
  • 2.12. Stem and cap height design2.13. Junction points; 2.14. Stroke ending; 2.15. Curve; 3. Comparing Arabic with Latin Font Characteristics; 3.1. Overview; 3.2. Feature of Arabic script compared to Latin script; 3.3. Anatomy of characters; 3.4. Design considerations; References; Chapter 2 Chinese Handwriting Database Building and Benchmarking; 1. Introduction; 2. Data Collection Settings; 2.1. Character sets; 2.2. Data collection; 3. Data Annotation; 3.1. Annotation of offline data; 3.2. Annotation of online data; 3.3. Data format; 4. Statistics of Datasets; 5. Recommendations of Usage
  • 5.1. Data partitioning5.2. Research scenarios; 5.2.1. Handwritten document segmentation; 5.2.2. Handwritten character recognition; 5.2.3. Text line recognition; 5.2.4. Handwritten document retrieval; 5.2.5. Writer adaptation; 5.2.6. Writer identi cation; 6. Preliminary Evaluation; 7. Competition Results; 8. Conclusion; References; Chapter 3 CNN Based Handwritten Character Recognition; 1. Introduction; 1.1. Development of deep learning; 1.2. CNN for image understanding; 1.3. Character recognition by CNN; 2. Overview of the CNN-Based Handwritten Character Recognition System
  • 2.1. The framework of the system2.2. Recognition beyond human; 3. Random Distortion for Sample Generation; 4. Training Tricks of CNN Model; 5. Model Scale and Input Image Size; 6. Multi-Model Voting; 7. Conclusion; References; Chapter 4 Online Handwritten Chinese Character Recognition: From a Bayesian Approach to Deep Learning; 1. Introduction; 2. Online HCCR under a Bayesian Statistical Learning Framework; 2.1. General framework for OHCCR; 2.2. Preprocessing; 2.2.1. Normalization; 2.2.2. Data augmentation with deformation transformation; 2.2.3. Imaginary stroke technique
  • 2.3. Feature extraction2.3.1. Region partition with the meshing technique; 2.3.2. 8-directional feature extraction; 2.3.3. Signature of path features; 2.4. Feature dimension reduction with LDA; 2.5. MQDF classifier; 3. OHCCR-CNN: An End-to-End Approach for OHCCR using Deep Convolutional Neural Networks; 3.1. Brief introduction to CNNs; 3.1.1. Convolutional layer; 3.1.2. Pooling layer; 3.1.3. Softmax layer and loss function; 3.1.4. Platforms for training CNNs; 3.2. Domain knowledge-enhanced DCNN for OHCCR; 3.3. Efficient training of CNN using a new algorithm named DropSample
Control code
ocn981975306
Dimensions
unknown
Extent
1 online resource
Form of item
online
Isbn
9789813143685
Note
World Scientific Publishing
Other physical details
illustrations
Specific material designation
remote
System control number
(OCoLC)981975306

Library Locations

    • InternetBorrow it
      Albany, Auckland, 0632, NZ
Processing Feedback ...