ViTCAE: ViT-based Class-conditioned Autoencoder | Read Paper on Bytez