This book constitutes the refereed proceedings of the 18th International Conference on Frontiers in Handwriting Recognition, ICFHR 2022, which took place in Hyderabad, India, during December 4-7, 2022.
The 36 full papers and 1 short paper presented in this volume were carefully reviewed and selected from 61 submissions. The contributions were organized in topical sections as follows: Historical Document Processing; Signature Verification and Writer Identification; Symbol and Graphics Recognition; Handwriting Recognition and Understanding; Handwriting Datasets and Synthetic Handwriting Generation; Document Analysis and Processing.
Historical Document Processing.- A Few Shot Multi-Representation
Approach for N-gram Spotting in Historical Manuscripts.- Text Edges Guided
Network for Historical Document Super Resolution.- CurT: End-to-End Text Line
Detection in Historical Documents with Transformers.- Date Recognition in
Historical Parish Records.- Improving Isolated Glyph Classification Task for
Palm leaf Manuscripts.- Signature Verification and Writer
Identification.- Impact of Type of Convolution Operation on Performance of
Convolutional Neural Networks for Online Signature Verification.- COMPOSV++:
Light Weight Online Signature Verification Framework through Compound Feature
Extraction and Few-shot Learning.- Finger-Touch Direction Feature Using a
Frequency Distribution in the Writer Verification Base on Finger-Writing of a
Simple Symbol.- Self-Supervised Vision Transformers with Data Augmentation
Strategies using Morphological Operations for Writer Retrieval.- EAU-Net: A
New Edge-Attention based U-Net for Nationality Identification.- Progressive
Multitask Learning Network for Online Chinese Signature Segmentation and
Recognition.- Symbol and Graphics Recognition.- Musigraph: Optical Music
Recognition through Object Detection and Graph Neural Network.- Combining CNN
and Transformer as Encoder to Improve End-to-end Handwritten Mathematical
Expression Recognition Accuracy.- A Vision Transformer based Scene Text
Recognizer with Multi-Grained Encoding and Decoding.- Spatial Attention and
Syntax Rule Enhanced Tree Decoder for Offline Handwritten Mathematical
Expression Recognition.- Handwriting Recognition and Understanding.- FPRNet:
End-to-end Full-page Recognition Model for Handwritten Chinese Essay.- Active
Transfer Learning for Handwriting Recognition.- Recognition-free Question
Answering on Handwritten Document Collections.- Handwriting recognition and
automatic scoring for descriptive answers in Japanese language tests.- A
Weighted Combination of Semantic and Syntactic Word Image Representations.-
Combining Self-Training and Minimal Annotations for Handwritten Word
Recognition.- Script-Level Word Sample Augmentation for Few-shot
Handwritten Text Recognition.- Towards understanding and improving
handwriting with AI.- ChaCo: Character Contrastive Learning for Handwritten
Text Recognition.- Enhancing Indic Handwritten Text Recognition using Global
Semantic Information.- Yi Characters Online Handwriting Recognition Models
Based on Recurrent Neural Network: RnnNet-Yi and
ParallelRnnNet-Yi.- Self-Attention Networks for Non-Recurrent Handwritten
Text Recognition.- An Efficient Prototype-based Model for Handwritten Text
Recognition with Multi-Loss Fusion.- Handwriting Datasets and Synthetic
Handwriting Generation.- Urdu Handwritten Ligature Generation using
Generative Adversarial Networks (GANs).- SCUT-CAB: A New Benchmark Dataset of
Ancient Chinese Books with Complex Layouts for Document Layout Analysis.- A
Benchmark Gurmukhi Handwritten Character Dataset: Acquisition, Compilation,
and Recognition.- Synthetic Data Generation for Semantic Segmentation of
Lecture Videos.- Generating synthetic styled Chu Nom characters.- UOHTD: Urdu
Offline Handwritten Text Dataset.- Document Analysis and Processing.- DAZeTD:
Deep Analysis of Zones in Torn Documents.- CNN-based Ruled Line Removal in
Handwritten Documents.- Complex Table Structure Recognition in the Wild using
Transformer and Identity Matrix-based Augmentation.
