Specialized Image Descriptors Adaptation for Generated Images Recognition

P. 278–284.

Aleksei Samarin, Aleksei Toropov, Kotenko E., Nazarenko A., Elena Mikhailova, Valentin Malykh, Alexander Savelev, Motyko A.

This study introduces an innovative method for recognizing automatically generated images by utilizing adapted descriptors specifically designed to analyze unique structural and morphological features characteristic of artificially created content. The methodology focuses on analyzing features inherent to image generation processes, ensuring the optimization of descriptors for identifying complex and subtle patterns associated with generative algorithms. The integration of these specialized descriptors not only enhances recognition accuracy but also enables the extraction of interpretable features that provide deeper insights into the key principles of artificial content creation. To validate the effectiveness of the proposed method, an annotated dataset was developed and used to test and compare performance against various classification algorithms, including deep learningbased neural networks. Experimental results demonstrated that the proposed approach achieves high efficiency. These findings underscore the significance of the proposed methodology in advancing automated systems for the analysis of generated content in areas such as authenticity verification, media security, marketing, and digital validation while maintaining high computational efficiency and interpretability of results.

Keywords: computer vision, image classification, generated image recognition, image descriptors, synthetic images

In book

Proceedings of the 37th Conference of Open Innovations Association FRUCT, Helsinki, Finland, 14-16 May 2025, Issue 1 (Full Papers)

Vol. 37. , Helsinki: FRUCT Oy, 2025.