🤖 AI Summary
Prior work on BMI estimation relies predominantly on facial or postural imagery, lacking exploration of handwriting as a potential physiological biomarker. Method: This study pioneers the investigation of correlations between handwritten English characters and BMI, proposing a handwriting-to-BMI mapping paradigm. Using a custom dataset of lowercase letter handwriting from 48 subjects, we design a lightweight, end-to-end CNN model for non-invasive BMI prediction. Contribution/Results: Our model achieves 99.92% classification accuracy—surpassing AlexNet (99.69%) and InceptionV3 (99.53%)—demonstrating that handwriting dynamics encode stable, discriminative, BMI-related biometric signals. This work bridges the gap between graphology and physiological parameter estimation, establishing handwriting as a viable, low-cost, behavior-based modality for population-scale health screening without specialized hardware or invasive procedures.
📝 Abstract
A person's Body Mass Index, or BMI, is the most widely used parameter for assessing their health. BMI is a crucial predictor of potential diseases that may arise at higher body fat levels because it is correlated with body fat. Conversely, a community's or an individual's nutritional status can be determined using the BMI. Although deep learning models are used in several studies to estimate BMI from face photos and other data, no previous research established a clear connection between deep learning techniques for handwriting analysis and BMI prediction. This article addresses this research gap with a deep learning approach to estimating BMI from handwritten characters by developing a convolutional neural network (CNN). A dataset containing samples from 48 people in lowercase English scripts is successfully captured for the BMI prediction task. The proposed CNN-based approach reports a commendable accuracy of 99.92%. Performance comparison with other popular CNN architectures reveals that AlexNet and InceptionV3 achieve the second and third-best performance, with the accuracy of 99.69% and 99.53%, respectively.