An Intelligent Method for Recognizing Real and Generated Ukrainian-Language Voices

The modern development of speech synthesis technologies and convolutional neural networks (CNN) creates new opportunities and challenges in the field of voice recognition. One of the key tasks is to ensure the security and authenticity of audio information, especially in the context of information warfare and disinformation campaigns. In this context, the ability to recognize and identify generated voices becomes particularly important. This study aims to develop an intelligent method for recognizing real and generated Ukrainian-language voices using convolutional neural networks. The proposed method is based on the careful collection and preparation of data, the development and training of a CNN model, and its validation on test samples. Special attention is paid to ensuring the balance of the dataset, which is crucial for preventing model bias and improving its ability to generalize to unknown data. The model shows high performance in detecting real and generated voices, making this approach useful for local applications, particularly in Ukraine.