Towards the Influence of Text Quantity on Writer Retrieval

TL;DR

This study examines how the amount of text affects writer retrieval accuracy, showing that deep learning methods outperform traditional features especially with limited text, maintaining high accuracy with at least four lines.

Contribution

It provides a comprehensive analysis of text quantity impact on writer retrieval, comparing handcrafted and deep learning approaches across multiple datasets.

Findings

Performance drops 20-30% with only one line of text.

Deep learning methods outperform handcrafted features in low-text scenarios.

At least four lines of text maintain over 90% of full-page retrieval accuracy.

Abstract

This paper investigates the task of writer retrieval, which identifies documents authored by the same individual within a dataset based on handwriting similarities. While existing datasets and methodologies primarily focus on page level retrieval, we explore the impact of text quantity on writer retrieval performance by evaluating line- and word level retrieval. We examine three state-of-the-art writer retrieval systems, including both handcrafted and deep learning-based approaches, and analyze their performance using varying amounts of text. Our experiments on the CVL and IAM dataset demonstrate that while performance decreases by 20-30% when only one line of text is used as query and gallery, retrieval accuracy remains above 90% of full-page performance when at least four lines are included. We further show that text-dependent retrieval can maintain strong performance in low-text scenarios. Our findings also highlight the limitations of handcrafted features in low-text scenarios, with deep learning-based methods like NetVLAD outperforming traditional VLAD encoding.