With the exponential growth of high-quality fake images in social networks and media, it is necessary to develop recognition algorithms for this type of content. One of the most common types of image and video editing consists of duplicating areas of the image, known as the copy-move technique. Traditional image processing approaches manually look for patterns related to the duplicated content, limiting their use in mass data classification. In contrast, approaches based on deep learning have shown better performance and promising results, but they present generalization problems with a high dependence on training data and the need for appropriate selection of hyperparameters. To overcome this, we propose two approaches that use deep learning, a model by a custom architecture and a model by transfer learning. In each case, the impact of the depth of the network is analyzed in terms of precision (P), recall (R) and F1 score. Additionally, the problem of generalization is addressed with images from eight different open access datasets. Finally, the models are compared in terms of evaluation metrics, and training and inference times. The model by transfer learning of VGG-16 achieves metrics about 10% higher than the model by a custom architecture, however, it requires approximately twice as much inference time as the latter.

Copy-Move Forgery Detection (CMFD) Using Deep Learning for Image and Video Forensics.

Due to the exponential growth of high-quality fake photos on social media and the Internet, it is critical to develop robust forgery detection tools. Traditional picture- and video-editing techniques include copying areas of the image, referred to as the copy-move approach. The standard image processing methods physically search for patterns relevant to the duplicated material, restricting the usage in enormous data categorization. On the contrary, while deep learning (DL) models have exhibited improved performance, they have significant generalization concerns because of their high reliance on training datasets and the requirement for good hyperparameter selection. With this in mind, this article provides an automated deep learning-based fusion model for detecting and localizing copy-move forgeries (DLFM-CMDFC). The proposed DLFM-CMDFC technique combines models of generative adversarial networks (GANs) and densely connected networks (DenseNets). The two outputs are combined in the DLFM-CMDFC technique to create a layer for encoding the input vectors with the initial layer of an extreme learning machine (ELM) classifier. Additionally, the ELM model's weight and bias values are optimally adjusted using the artificial fish swarm algorithm (AFSA). The networks' outputs are supplied into the merger unit as input. Finally, a faked image is used to identify the difference between the input and target areas. Two benchmark datasets are used to validate the proposed model's performance. The experimental results established the proposed model's superiority over recently developed approaches.

/pdf/design-of-automated-deep-learning-based-fusion-model-for-2s3mmmhx.pdf

G. Sudhavani

Papers

An efficient copy move forgery detection using adaptive watershed segmentation with AGSO and hybrid feature extraction