VolumeGAN: A Generative Adversarial Network for High-fidelity 3D-aware Image Synthesis - 42Papers