A Conformer Based Automated Speech Recognition for Armenian Language

The article is aimed to represent Armenian automated speech recognition model and its applications in different fields of economy. Because the lack corpora, this we fine-tuned voice text symbol generating parts using a Conformer pre-trained compact language model. focuses attention readers on problem recognizing human transforming it text, especially for non-mainstream languages. paper prepared with scientific abstraction combined analysis many recent implementations discussed approach. sources' credibility, relevance authenticity have been confirmed by their extensive research. As conclude, though pretty challenging develop ASR languages, was proven that employment Conformerbased transformers conjunction models effective recognition. It also technique employed applicable other languages too, some adjustments.