Automated gender classification from handwriting: a systematic survey

Automatically identifying the gender of a writer from a handwritten sample is an essential task in various domains, including historical document analysis, handwriting biometrics, and psychology. Technological advances in computer vision and image analysis have yielded various techniques suitable for this task, each with its own merits and limitations. However, a systematic survey of these techniques, which would provide researchers guidance on selecting the appropriate approach, is currently lacking. To address this gap, we used a predefined query to select and then analyze a selection of peer-reviewed studies published between 2012 and 2021 that presented automatic methods for the classification of gender from handwriting. Next, we describe and categorize the feature extraction methods applied and classifiers used in each study, overview the existing datasets, and compare the results across studies. Finally, based on these data, we specify yet-unresolved issues in the field and provide recommendations for the development of new and improved classification methods.