Improving the Training of Data-Efficient GANs via Quality Aware Dynamic Discriminator Rejection Sampling | Read Paper on Bytez