A Novel CNN-BiLSTM Fusion Model for Robust Age Estimation from Facial Images

This study unveils a novel approach to age estimation from facial images, achieved by amalgamating Convolutional Neural Networks (CNNs) and Bidirectional Long Short-Term Memory networks (BiLSTMs) into a cohesive fusion model. A dataset, comprised of facial photographs of 80 individuals aged between 16 and 25 years, was meticulously assembled to facilitate the development of a robust and reliable age estimation model. This specific age bracket is characterized by significant facial transformations attributable to both growth and ageing, thereby posing a considerable challenge for accurate age estimation. The introduced model employs metric regression techniques within a structure that integrates CNN and BiLSTM, capitalizing on their synergistic capabilities to estimate age with precision. CNNs are renowned for their prowess in feature extraction, while BiLSTMs excel in discerning temporal relationships. This amalgamation ensures the comprehensive capture of both static and dynamic facial signals instrumental for precise age estimation. Evaluative metrics including Mean Absolute Error (MAE), Mean Squared Error (MSE), and Root Mean Squared Error (RMSE) are employed to gauge the model’s robustness and efficacy within the designated age bracket, offering insights into its competitive standing in the realm of age estimation from facial imagery.