Current Volume 9
Abstract
Deepfake content has rapidly emerged as a major threat across digital media platforms, enabling the creation of highly realistic manipulated images that can mislead viewers, spread misinformation, and compromise security systems. Traditional image forensics approaches struggle to distinguish such sophisticated forgeries, especially when manipulation artifacts are subtle or deliberately concealed. To address this challenge, this work presents a robust deepfake image detection framework that integrates Error Level Analysis (ELA) with the ResNet18 deep neural network using transfer learning. ELA serves as a forensic preprocessing step that highlights compression inconsistencies introduced during image tampering. By recompressing the input image at a known JPEG quality and computing the pixel-level difference between the original and recompressed versions, ELA produces a high-frequency residual map that exposes hidden manipulation artifacts not visible in raw images. These ELA- generated images are then used to train and evaluate a modified ResNet18 model, where the final fully connected layer is replaced to accommodate binary classification of real and fake images. The proposed system is trained and tested on the publicly avail- able Deepfake and Real Images Dataset from Kaggle, comprising manipulated and pristine images across diverse categories. Experimental results demonstrate that the ELA-ResNet18 com- bination significantly enhances detection performance, achieving an overall classification accuracy of 97
Citations
IRE Journals:
Dr. Nagul Meera Sayyed, Pokala Venkat, Rowthu Babu Nagendra Kumar, Shaik Sai Raheem "Deepfake Image Detection Using Error Level Analysis and ResNet18 Transfer Learning" Iconic Research And Engineering Journals Volume 9 Issue 10 2026 Page 3266-3273 https://doi.org/10.64388/IREV9I10-1717046
IEEE:
Dr. Nagul Meera Sayyed, Pokala Venkat, Rowthu Babu Nagendra Kumar, Shaik Sai Raheem
"Deepfake Image Detection Using Error Level Analysis and ResNet18 Transfer Learning" Iconic Research And Engineering Journals, 9(10) https://doi.org/10.64388/IREV9I10-1717046
About IRE Journals
IRE Journals is an open access online journals established with an aim to publish high quality of research work in various diciplines. We have more than 6 years in publications. Over 2500+ papers already published.
Important Links
For Authors
Volume 8:
Issue 1, Issue 2, Issue 3, Issue 4, Issue 5, Issue 6, Issue 7, Issue 8, Issue 9, Issue 10, Issue 11, Issue 12Volume 7:
Issue 1, Issue 2, Issue 3, Issue 4, Issue 5, Issue 6, Issue 7, Issue 8, Issue 9, Issue 10, Issue 11, Issue 12Volume 6:
Issue 1, Issue 2, Issue 3, Issue 4, Issue 5, Issue 6, Issue 7, Issue 8, Issue 9, Issue 10, Issue 11, Issue 12Volume 5:
Issue 1, Issue 2, Issue 3, Issue 4, Issue 5, Issue 6, Issue 7, Issue 8, Issue 9, Issue 10, Issue 11, Issue 12Volume 4:
Issue 1, Issue 2, Issue 3, Issue 4, Issue 5, Issue 6, Issue 7, Issue 8, Issue 9, Issue 10, Issue 11, Issue 12Volume 3:
Issue 1, Issue 2, Issue 3, Issue 4, Issue 5, Issue 6, Issue 7, Issue 8, Issue 9, Issue 10, Issue 11, Issue 12