A group experiment to find on technical paper and one patent in signal processing on a 1D signal . The application that we selected was 'Noise Reduction Using Adaptive Filters'. The group members are Kapil Rawal, Prerana Sarode,Chinmay Upadhyae,vaishnav dandge and Harshit Shukla.
Summary
A new adptive algorithm named affine projection is used rather than traditional methods for noise cancelation like LMS and NLMS ,the output is compared on the basis of signal-to-noise ratio.The result showed that affine projection algorithm gave high signal to noise ratio then LMS and NLMS.
I think we could use Mel-Frequency Cepstrum Coefficient for further reducing the noise and proper detection of input voice.
ReplyDeletedon't know about detection by mfcc but they are not very robust in the presence of additive noise,thus modifications to the basic MFCC algorithm to improve robustness by raising the log-mel-amplitudes to a suitable power.
DeleteMan , the Affine Projection algorithm was difficult to understand let alone implement
ReplyDeleteRight, had difficulty to implement it . your help on this topic will be appreciated.
Delete