A Transformer Based Handwriting Recognition System Jointly Using Online and Offline Features

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lodh, Ayush, Chakraborty, Ritabrata, Palaiahnakote, Shivakumara, Pal, Umapada
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916810772709376
author Lodh, Ayush
Chakraborty, Ritabrata
Palaiahnakote, Shivakumara
Pal, Umapada
author_facet Lodh, Ayush
Chakraborty, Ritabrata
Palaiahnakote, Shivakumara
Pal, Umapada
contents We posit that handwriting recognition benefits from complementary cues carried by the rasterized complex glyph and the pen's trajectory, yet most systems exploit only one modality. We introduce an end-to-end network that performs early fusion of offline images and online stroke data within a shared latent space. A patch encoder converts the grayscale crop into fixed-length visual tokens, while a lightweight transformer embeds the $(x, y, \text{pen})$ sequence. Learnable latent queries attend jointly to both token streams, yielding context-enhanced stroke embeddings that are pooled and decoded under a cross-entropy loss objective. Because integration occurs before any high-level classification, temporal cues reinforce each other during representation learning, producing stronger writer independence. Comprehensive experiments on IAMOn-DB and VNOn-DB demonstrate that our approach achieves state-of-the-art accuracy, exceeding previous bests by up to 1\%. Our study also shows adaptation of this pipeline with gesturification on the ISI-Air dataset. Our code can be found here.
format Preprint
id arxiv_https___arxiv_org_abs_2506_20255
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle A Transformer Based Handwriting Recognition System Jointly Using Online and Offline Features
Lodh, Ayush
Chakraborty, Ritabrata
Palaiahnakote, Shivakumara
Pal, Umapada
Computer Vision and Pattern Recognition
Machine Learning
We posit that handwriting recognition benefits from complementary cues carried by the rasterized complex glyph and the pen's trajectory, yet most systems exploit only one modality. We introduce an end-to-end network that performs early fusion of offline images and online stroke data within a shared latent space. A patch encoder converts the grayscale crop into fixed-length visual tokens, while a lightweight transformer embeds the $(x, y, \text{pen})$ sequence. Learnable latent queries attend jointly to both token streams, yielding context-enhanced stroke embeddings that are pooled and decoded under a cross-entropy loss objective. Because integration occurs before any high-level classification, temporal cues reinforce each other during representation learning, producing stronger writer independence. Comprehensive experiments on IAMOn-DB and VNOn-DB demonstrate that our approach achieves state-of-the-art accuracy, exceeding previous bests by up to 1\%. Our study also shows adaptation of this pipeline with gesturification on the ISI-Air dataset. Our code can be found here.
title A Transformer Based Handwriting Recognition System Jointly Using Online and Offline Features
topic Computer Vision and Pattern Recognition
Machine Learning
url https://arxiv.org/abs/2506.20255