Assume that an image vector x = Ψs. s is a sparse vector in which the image vector x of length N x 1 is sparse in the wavelet Ψ basis.

I have issue in finding the measurement matrix A= φ Ψ where φ is the random binary/Gaussian matrix of size K x N and Ψ is the wavelet basis of size N x N. May I know what kind of wavelet that is good in reconstructing using L1 norm (e.g lasso)? My question is how to build the measurement matrix in Python/MATLAB?

  • $\begingroup$ I do not understand your question "how to build the measurement matrix in Python/MATLAB?" simply generate a KxN random matrix $\endgroup$
    – MimSaad
    Commented Jun 14, 2021 at 9:19
  • $\begingroup$ @MimSaad Hi. I am interested to train a network with weights of {-1, 1} in order to simplify the encoding process. Do you have any good idea for this? $\endgroup$
    – Yvon
    Commented Jul 31, 2021 at 7:06


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.