OCR-VQGAN: Taming Text-within-Image Generation | IEEE Conference Publication | IEEE Xplore