Somayeh Hosseini MEE10:73, pp. 73. ING/School of Engineering, 2010.
This project aims at proposing a new noise reduction technique for speech
recognition purposes. The proposed method called mapping based noise
reduction is performed on the feature vectors extracted from speech signals.
In this work the dimensionality reduction functionality of algorithms
such as Locally Linear Embedding and Principal Component Analysis is
exploited to map the corrupted speech feature vectors to their corresponding
noise-free feature vectors. The feature vectors are first mapped to the
lower dimensional space and in this space the nearest clean vector to each
noisy vector is found, mapped back again to the original space and given
as the input to the speech recognition system. This approach is examined
on the speech signals with artificially added wind noise with different signal
to noise ratio values and articulated by two different speakers.