SINE: SINgle Image Editing With Text-to-Image Diffusion Models | Read Paper on Bytez