Robust uncertainty principles: exact signal reconstruction from highly incomplete frequency information
read more
Citations
Compressed sensing
Distributed Optimization and Statistical Learning Via the Alternating Direction Method of Multipliers
An Introduction To Compressive Sampling
Signal Recovery From Random Measurements Via Orthogonal Matching Pursuit
Decoding by linear programming
References
Numerical Optimization
Nonlinear total variation based noise removal algorithms
Atomic Decomposition by Basis Pursuit
Greed is good: algorithmic results for sparse approximation
Related Papers (5)
Frequently Asked Questions (11)
Q2. What is the commonly discussed approach in the literature for reconstructing an object from polar frequency?
Frequently discussed approaches in the literature of medical imaging for reconstructing an object from polar frequency samples are the so-called filtered backprojection algorithms.
Q3. Why is the distribution of the eigenvalues of random variables in (2.16) independent?
Because the random variables in (2.16) are independent andhave all the same distribution, the quantity depends only on the equivalence relation and not on the value of itself.
Q4. How does the proof of the theorem work?
The proof of the theorem then proceeds by developing a recursive inequality on the central term in this second expansion, which is done in Section IV-E.
Q5. What is the conjecture that is essentially the content of Section I-G?
The earlier [4] also contains a conjecture that more powerful uncertainty principles may exist if one of , is chosen at random, which is essentially the content of Section I-G here.
Q6. What is the distribution of the eigenvalues of such operators?
The distribution of the eigenvalues of such operators was studied by Landau and others [27]–[29] while developing the prolate spheroidal wave functions that are now commonly used in signal processing and communications.
Q7. What is the way to recover from a convex matrix?
To the best of their knowledge, one would essentially need to let vary over all subsets of cardinality , checking for each one whether is in the range of or not, and then invert the relevant minor of the Fourier matrix to recover once is determined.
Q8. What is the simplest way to find a trigonometric polynomial?
The following lemma shows that a necessary and sufficient condition for the solution to be the solution to is the existence of a trigonometric polynomial whose Fourier transform is supported on , matches on , and has magnitude strictly less than elsewhere.
Q9. Why is the paper unable to elaborate on this fact?
Because of space limitation, the authors are unable to elaborate on this fact and its implications further, but will do so in a companion paper.
Q10. What is the function that the authors exchanged?
(4.43)Note that the authors voluntarily exchanged the function arguments to reflect the idea that the authors shall view as a function of while will serve as a parameter.
Q11. What is the simplest way to recover a signal?
Theorem 1.3 states that for any signal supported on an arbitrary set in the time domain, recovers exactly—with high probability— from a number of frequency samples that is within a constant of .