Ziming Qi
Doctor of Philosophy, (Computer Engineering)
Study Completed: 2009
College of Sciences
Citation
Thesis Title
Real-time adaptive noicse cancellation for automatic speech recognition in a car environment
Read article at Massey Research Online: 
Mr Qi proposed a robust method for improved performance of real-time speech enhancement and noise cancellation in a car environment. He presented a novel method to simplify the use of microphones and reduce the real-time environmental noise in order to control devices using Automatic Speech Recognition (ASR). In order to cancel the unwanted voice in a car, Mr Qi proposed an Adaptive Noise Cancellation (ANC), which uses a system with an acoustic beamformer based Voice Activity Detector (VAD), Normalize least mean squares Adaptive filter and an Adaptive Wiener Filter. Mr Q applied an ASR algorithm to test the quality of this hybrid ANC. Mr Qi’s research produced improved performance in speech enhancement and noise cancellation with several favourable effects, namely: a vast improvement in signal-to-noise ratio and the ability to reject the effect of interference from other speakers.
Supervisors
Dr Tom Moir
Dr Mohammad Abdurrashid