Data Sufficiency for Online Writer Identification: A Comparative Study of Writer-Style Space vs. Feature Space Models

Abstract

A key factor in building effective writer identification/verification systems is the amount of data required to build the underlying models. In this research we systematically examine data sufficiency bounds for two broad approaches to online writer identification -feature space models vs. writer-style space models. We report results from 40 experiments conducted on two publicly available datasets and also test identification performance for the target models using two different feature functions. Our findings show that the writer-style space model gives higher identification performance for a given level of data and further, achieves high performance levels with lesser data costs. This model appears to require as less as 20 words per page to achieve identification performance close to 80% and reaches more than 90% accuracy with higher levels of data enrollment.

IAMonSense: multi-level handwriting classification using spatiotemporal information

“…Shivram et al [32] used the IAM-onDB dataset for writer identification using hierarchical Bayesian models based on their online handwriting data. In another method [33], Shivram et al presented a comparative study for online writer identification to evaluate the influence of writing style (memetic factors) and an individual's handwriting features. Ahmed et al [34] used the IAM-onDB dataset to implement an ensemble-based classifier system to predict the gender of the writer based on his handwriting data.…”

Improving Arabic writer identification using score‐level fusion of textural descriptors

“…While, text‐dependent methods require writing samples with same textual content for comparison, text‐independent methods allow identification of writers independent of the semantic content. Likewise, as a function of handwriting acquisition method, identification methods are distinguished into online [31, 32] and offline [3, 4, 30, 33] techniques. Offline methods employ digitised images of handwriting and rely on statistical or structural features extracted locally or globally from the handwriting images.…”