A Comparison of Machine Learning Techniques for the Identification of Real and AIGenerated Images

Read full text Summarise with VIRAT
Authors: Lokeswari Ekambaram, Virupakshi Lahya
Cited by: 0
References: 1

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Differentiating between real and synthetic images has become more challenging due to the growing usage of artificial intelligence in image production. Highly realistic graphics can be produced by sophisticated models like Generative Adversarial Networks and diffusion-based approaches, which raises questions regarding content authenticity, digital fraud, and disinformation. This research uses machine learning and deep learning techniques to compare several detection algorithms for recognising real and AI-generated photos. To ascertain the authenticity of photographs, the suggested method preprocesses image data, extracts pertinent features, and applies many classification models. Performance metrics, including accuracy, precision, recall, and F1-score, are used to assess a variety of methods, including convolutional neural networks and conventional machine learning classifiers. The study outlines the advantages and disadvantages of each technique for identifying minute distinctions between actual and artificial images. According to experimental findings, deep learning models perform better than conventional methods in identifying intricate patterns and discrepancies in AI-generated photos. The results of this study help to strengthen trust in digital media platforms and automated picture verification systems. The suggested method can be expanded to identify sophisticated deepfake content and enhance practical security applications.

Fields of Study & Subjects

Digital Media Forensic Detection Generative Adversarial Networks and Image Synthesis Misinformation and Its Impacts

Publication details

Journal / source
JOURNAL OF EMERGING TRENDS AND NOVEL RESEARCH
Volume
4
Issue
3
Publication date
1 January 2026
Language
EN