General Adversarial Network - What is the major problem with GAN?

General Adversarial Networks' (GANs) main flaw is that they can't focus on the full distribution of data. Because the GANs' goal function is comparable to the Jensen-Shannon divergence (JSD), this is the result. Minimizing JSD only produces a decent fit to the primary mode and does not provide high-quality pictures, according to studies, even for a bi-model distribution.