# Predicting Movie Production Years through Facial Recognition of Actors with Machine Learning

arXiv: https://arxiv.org/abs/2504.01047 · HTML: https://mlsanity.com/paper/2504.01047/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 9 of its 64 references that are in the map.
- First posted on arXiv 1 Apr 2025.

## Builds on (most cited first, 9 of 9)

- [Very Deep Convolutional Networks for Large-Scale Image Recognition](https://mlsanity.com/paper/1409.1556/) (2014)
- [ImageNet Large Scale Visual Recognition Challenge](https://mlsanity.com/paper/1409.0575/) (2014)
- [FaceNet: A Unified Embedding for Face Recognition and Clustering](https://mlsanity.com/paper/1503.03832/) (2015)
- [SqueezeNet: AlexNet-level accuracy with 50x fewer parameters and <0.5MB model size](https://mlsanity.com/paper/1602.07360/) (2016)
- [A Neural Representation of Sketch Drawings](https://mlsanity.com/paper/1704.03477/) (2017)
- [Adversarial Examples Improve Image Recognition](https://mlsanity.com/paper/1911.09665/) (2019)
- [Towards Fairer Datasets: Filtering and Balancing the Distribution of the People Subtree in the ImageNet Hierarchy](https://mlsanity.com/paper/1912.07726/) (2019)
- [Learning Robust Visual-Semantic Embeddings](https://mlsanity.com/paper/1703.05908/) (2017)
- [Eight Years of Face Recognition Research: Reproducibility, Achievements and Open Issues](https://mlsanity.com/paper/2208.04040/) (2022)

## Abstract (arXiv)

This study used machine learning algorithms to identify actors and extract the age of actors from images taken randomly from movies. The use of images taken from Arab movies includes challenges such as non-uniform lighting, different and multiple poses for the actors and multiple elements with the actor or a group of actors. Additionally, the use of make-up, wigs, beards, and wearing different accessories and costumes made it difficult for the system to identify the personality of the same actor. The Arab Actors Dataset-AAD comprises 574 images sourced from various movies, encompassing both black and white as well as color compositions. The images depict complete scenes or fragments thereof. Multiple models were employed for feature extraction, and diverse machine learning algorithms were utilized during the classification and prediction stages to determine the most effective algorithm for handling such image types. The study demonstrated the effectiveness of the Logistic Regression model exhibited the best performance compared to other models in the training phase, as evidenced by its AUC, precision, CA and F1score values of 99%, 86%, 85.5% and 84.2% respectively. The findings of this study can be used to improve the precision and reliability of facial recognition technology for various uses as with movies search services, movie suggestion algorithms, and genre classification of movies.