favorite16In this paper, a fast iterative algorithm is proposed for recovering spectrally sparse signals whose frequencies can be any values in the continuous domain [0, 1) from a small amount of time domain samples.
favorite13We propose an efficient algorithm based on projected Wirtinger gradient descent for this particular spectral signal recovery problem.
favorite2To efficiently recover high-dimensional signals, this paper proposes a projected Wirtinger gradient descent (PWGD) method for low-rank Hankel matrix completion.
favorite0Though robust signal recovery is guaranteed theoretically through these methods in [11, 27, 14], convex optimization based low-rank structured matrix completions are not computationally efficient- the resulting optimization problems contain O(N 2 ) unknowns explicitly, where N is the dimension of signal.
favorite21After converting the spectrally sparse signal recovery into a low rank structured matrix completion problem, we propose an efficient feasible point approach, named projected Wirtinger gradient descent (PWGD) algorithm, to efficiently solve this structured matrix completion problem.
favorite4Projected Wirtinger Gradient Descent for Spectral Compressed Sensing Jian-Feng Cai. Abstract This paper considers reconstructing a spectrally sparse signal from a small number of randomly observed time-domain samples.