Speech Recognition Utilizing Adaptive Immune Clonal Neural Networks

• Nan Jiang,
• Jia Qin,
• Suyuan Li,
• Yongheng Pang,

Vol. 19, No. 9, September 30, 2025
10.3837/tiis.2025.09.014, Download Paper (Free):

• speech recognition
• Adaptive optimization
• Immune Clonal
• Neural Networks
• real-time systems
• Noise-robust recognition

Abstract

In the field of practical speech recognition systems, there is an urgent need for real-time and high-fidelity recognition capabilities even in the noisy environment of daily ambient noise. The adaptive immune clonal neural network is a new speech recognition algorithm that combines biological immune mechanisms with the dynamic optimization of neural networks. This method has a dynamic network structure adjustment mechanism, which can dynamically adjust the network structure according to the complexity of the input speech signal, and achieve collaborative optimization of parameters and structure through the immune mechanism, strengthening the extraction ability of key acoustic features. Different from conventional algorithms, the adaptive immune clonal neural network introduces an adaptive mutation rate, effectively improving the convergence speed and reducing the number of training iterations. At the same time, it has the characteristics of a robust immune memory mechanism and can quickly recognize the speech instructions of specific speakers. The average speech recognition accuracy rate of this method is 89.25%. It shows significant advantages in dimensions such as dynamic adaptability and noise robustness, breaking through the static limitations of traditional models and is especially suitable for speech recognition tasks in complex scenarios.

Statistics

Cite this article