Sieť typu Transformer a jej použitie

Abstract

{The Transformer network is a deep neural network architecture that has recently demonstrated success in natural language processing (NLP) and other tasks due to its ability to model long-range dependencies, improve parallel processing, and reduce computational complexity. It has been successfully used in many NLP tasks and has demonstrated superior performance compared to traditional models such as recurrent and convolutional neural networks. One of the areas in which the Transformer network has also made a difference is automatic speech recognition (ASR). ASR is a challenging task due to the variability and complexity of human speech. This paper explores the potential of the Transformer network for spoken word-to-text transcription and investigates its effectiveness on Slovak speech data in comparison with recurrent ASR models.