Monday, 13 March 2017

Experiment one : Discrete Convolution and Correlation


Part one: Convolution

Convolution is used to find output of the system.

We calculated 

  1. Linear convolution
  2. Circular convolution
  3. Linear convolution using circular convolution.

Learnings from experiment:

  • In linear convolution, length of output signal is one less than the addition of length of both input signals.
  • In circular convolution:
1) Length of output signal is same as that of input signal with higher length.
2) Last few values are aliased on initial few values. Hence we get aliased output.


Part two: Correlation
Correlation is used to find the degree of similarity between two signals.


We calculated 
  1. Autocorrelation
  2. Cross-correlation
Learnings from experiment:
  • Autocorrelation is an even signal.
  • Autocorrelation of delayed input signal is same as that of autocorrelation of original input signal.
  • Cross-correlation of input signal with delayed input signal is same as advanced autocorrelated input signal.
  • In cross-correlation of scaled input signal, output is scaled by same factor as that of input signal.

9 comments:

  1. Very systematic explanation.

    ReplyDelete
  2. correlation gives the similarity between two signals.

    ReplyDelete
    Replies
    1. Yes. It compares both signals and gives degree of similarity between them.

      Delete
    2. Auto correlation results give energy of the signal at 0th value

      Delete
  3. Circular convolution gives aliased output.

    ReplyDelete
    Replies
    1. Yes,circular convolution of two finite-length sequences is equivalent to the linear convolution of the two sequences, followed by time aliasing.

      Delete

Paper Review: Comparison of Parametric Representations for Monosyllabic Word Recognition in Continuously Spoken Sentences

Title: Comparison of Parametric Representations for Monosyllabic Word Recognition in Continuously Spoken Sentences Author: Steven B. Davi...