Transfer Learning-Based Deep Residual Learning for Speech Recognition in Clean and Noisy Environments

Abstract

Addressing the detrimental impact of non-stationary environmental noise on automatic speech recognition (ASR) has been a persistent and significant research focus. Despite advancements, this challenge continues to be a major concern. Recently, data-driven supervised approaches, such as deep neural networks, have emerged as promising alternatives to traditional unsupervised methods. With extensive training, these approaches have the potential to overcome the challenges posed by diverse real-life acoustic environments. In this light, this paper introduces a novel neural framework that incorporates a robust frontend into ASR systems in both clean and noisy environments. Utilizing the Aurora-2 speech database, the authors evaluate the effectiveness of an acoustic feature set for Mel-frequency, employing the approach of transfer learning based on Residual neural network (ResNet). The experimental results demonstrate a significant improvement in recognition accuracy compared to convolutional neural networks (CNN) and long short-term memory (LSTM) networks. They achieved accuracies of 98.94% in clean and 91.21% in noisy mode.

References in corpus (6)

• Automatic Speech Recognition using Advanced Deep Learning Approaches: A survey
• Deep Transfer Learning for Automatic Speech Recognition: Towards Better Generalization
• Transformers and Large Language Models for Efficient Intrusion Detection Systems: A Comprehensive Survey
• Artificial Intelligence for Cochlear Implants: Review of Strategies, Challenges, and Perspectives
• Automatic Speech Recognition with BERT and CTC Transformers: A Review
• Machine Learning and Transformers for Thyroid Carcinoma Diagnosis: A Review