Порівняння систем розпізнавання голосу основаних на статистичній моделі і глибокі нейронні мережі

Abstract

This work provides an overview of different approaches to voice recognition as statistical models like Hidden Markov Models and deep neural networks like CNNs, RNNs, and LSTMs withthe challenges of noise in voice recognition, and the benefits of DNN-based models in handling noisy environments.