Source Camera Model Identification Using Deep Features and Ensemble Learning
Rukmini Maiti, Vibhav Prakash Singh · 2025
Camera model identification refers to, finding the source camera model used while capturing an image. This is an important research area in digital forensics. This capability is essential to verify the authenticity of digital content, investigate cybercrimes, and protect intellectual property. In this study, we explore advanced methods for camera model identification using a hybrid approach. The model integrates deep learning techniques, which uses several layers of artificial neural network to learn features from given data. Features are extracted using advanced deep learning architectures: ResNet and DenseNet. To refine the achieved results, the extracted features are classified using machine learning algorithms such as Support Vector Machines (SVM),$k$-Nearest Neighbors ($k$-NN), and Random Forest. This research highlights the synergy of combining deep learning and machine learning techniques for camera model identification and sets the stage for future advancements in digital forensics.