Alternating direction method of multipliers with compound gaussian and generative adversarialnetwork priors

A vehicle, the vehicle comprising:

• a data acquisition system, comprising:
• a radar configured to:
• transmit a waveform towards a target, and
• receive a signal from the target, wherein the signal is a reflected version of the waveform;
• and
• a computing device, comprising:
• a memory,
• a display, and
• a processor, wherein the processor is configured to perform operations comprising:
• determining, based on the signal, a compound gaussian (CG) prior and a generative adversarial network (GAN) prior, wherein G is the generative neural network (NN) from the GAN and px is low-dimensional latent distribution for G,
• determining, based on the CG prior, the GAN prior, the signal, a plurality of initialization parameters, including x, z, u, c, φ2 and φ1, wherein z represents a scale variable, u represents a Gaussian variable, c represents input model parameters, φ2 and φ1 represent dual variable estimates for an iteration k, and x represents a latent vector x~px such that c=G(x),
• performing image estimation for the signal based on the initialization parameters until a predetermined tolerance parameter τ is reached that ensures that a mean squared error in both a feasibility gap and a change in each of the variables x, z, u, and c is sufficiently small, and
• sending output resulting from the image estimation to the display.

A method, comprising:

• transmitting, using a radar device, a waveform towards a target;
• receiving, using the radar device, a signal from the target, wherein the signal is a reflected version of the waveform;
• determining, based on the signal, a compound gaussian (CG) prior and a generative adversarial network (GAN) prior, wherein G is the generative neural network (NN) from the GAN and px is low-dimensional latent distribution for G;
• determining, based on the CG prior, the GAN prior, the signal, a plurality of initialization parameters, including x, z, u, c, φ2 and φ1, wherein z represents a scale variable, u represents a Gaussian variable, c represents input model parameters, φ2 and φ1 represent dual variable estimates for an iteration k, and x represents a latent vector x~px such that c=G(x);
• and
• performing image estimation for the signal based on the initialization parameters until a predetermined tolerance parameter t is reached that ensures that a mean squared error in both a feasibility gap and a change in each of the variables x, z, u, and c is sufficiently small.