Detecting Fake Images Using Neuro-Fuzzy Inference Systems: A Brief Comparative Analysis
Tayasan Milinda H. Gedara, Vincenzo Loia, Stefania Tomasiello · 2024
In an era where image manipulation is easily accessible, detecting digital image forgery has become more challenging. The manuscript delves into the essential issue of picture forgery and discusses the use of the standard Adaptive Neuro-Fuzzy Inference System (ANFIS) and some of its variants for identifying and analyzing altered digital image content. Although ANFIS is a well-known technique, its use for such problems has not been explored so far, to the best of our knowledge. The proposed methodology uses feature extraction to get the numerical input to ANFIS. Unlike deep learning techniques, most ANFIS schemes have the advantage of being interpretable, without requiring a long training time and ensuring a good performance. Compared to state-of-the-art techniques in the deep learning realm, the proposed method turned out to work well with several forgery datasets. It offers a viable solution for real-time detection, paving the way for future applications in digital content authentication.