Skip to Content
Doctor of Philosophy, (Engineering)
Study Completed: 2014
College of Sciences
Multi -microphone Speech Enhancement Technique using a Novel Neural Network Beamformer
Speech based applications in real world noisy environments are still somewhat limited. Ms Yoganathan presented a novel speech enhancement algorithm, in order to improve the performance of these systems by reducing the background noises from the acquired speech signal. She investigated the possibility of applying nonlinear neural network based filters to the noise reduction area of the switching adaptive filtering structure. Her algorithm showed promising noise reduction improvement over the previous linear structures. Most of the work in this area only concentrates on dual-microphone systems and her research explores a four channel microphone beamformer using the neural ntework noise canceller. In addition, the basic idea of using speech to control the vibration monitoring system under noisy environment was explored. By using the proposed neural network based beamformer with the speech recognition system, the recognition rate of the system was significantly improved.
Page authorised by Web Content Manager
Last updated on Tuesday 04 April 2017