ДОСЛІДЖЕННЯ МЕТОДІВ ТА МОДЕЛЕЙ АВТОМАТИЧНОГО РОЗПІЗНАВАННЯУЧАСНИКІВ АУДІО РОЗМОВИ

The object of the study is the process of automatic speech recognition (ASR) and one of the branch of it – speaker recognition or diarization (SD).
The purpose of the work is to study the stages of the process of speech and speakers recognition, the analysis of methods and models of machine learning and neural networks, as well as modern frameworks for training speech recognition models and diarization. Based on the received results of the study it would be selected the existing models, systems or products that would play a role of starting point in the creation of a custom ASR and SD models for
enhancing those processes and bringing more values and benefits in different areas of human being.