Survey on Image Immunizer an Image Tamper Resilient Multi Task Learning Scheme for Image Lossesless Auto- Recovery

The development of image tampering and forgery detection research accelerated during the rise of artificial intelligence and deep learning technology. The research field has developed multiple detection and localization systems for tampered images which originated from traditional image processing along with contemporary deep learning model development. Various contemporary methods for face inpainting coupled with image forgery detection strategies are studied within this research through GANs and CNNs along with deep learning classification and watermarking schemes. The study explores modern detection patterns using YOLOv5 as well as CNN-based classification for false image analysis through semantic segmentation for tampering identification. The research analyzes noise convolution networks as well as multi-scale residual deep CNNs and dual-domain convolutional neural networks. The methods demonstrate success in maintaining image authenticity and security and they enable new opportunities for designing better detection systems.