1 paper
Minghui Zhang, Jinxin Shi, Yifan Chang +12
Text-to-image generation has reached photorealistic quality, yet state-of-the-art systems remain unreliable at producing scientific diagrams, whose value depends not on appearance…