Authors - Su Thet Oo, Ah Nge Htwe, Nilar Aye Abstract - The automatic detection of AI-generated art images is essential for distinguishing authentic human creations from artificial ones. This process is critical for authenticity verification, provenance control, misinformation management, and digital forensics. With the rapid evolution of deep learning content generation, the existing detection approaches within artistic imagery remain an underexplored domain characterized by artworks that differ widely in style and often contain non-standard, complex, or distorted visual patterns. The proposed model is an empirical study of a fine-tuned CNN-based generative art detection to classify real and human-created art accurately by learning discriminative visual features such as texture, structure, and statistical patterns, adapting a pre-trained CNN model and also finetuning architecture layers and defining the spatial dimension, which is used to determine the level of detail captured in feature extraction and classification. In our system, utilizing a balanced dataset consisting of real and AI-generated art images, the system was trained and evaluated, where a base VGG16 net in traditional architecture and this architecture of pre-trained and fine-tuned VGG16 with hyperparameter tuning of task-specific input representation and data augmentation, layer optimization strategies using the same balanced dataset, with results benchmarked against a strong baseline.
Open Zoom