An idiom says, “read between the lines”, which means trying to understand someone's real feelings or intentions from what they say or write. The AI model generates images based on given texts in the text-to-image artificial intelligence context. In this project, I am surprised that when I use my poems to generate images, the default outputs of Stable Diffusion v1.5 are always like photos. I view this as a way to take photos between the lines, letting AI reveal its “understanding” of my poems. Thus, I choose the name “Pictures Between The Lines”.
The original poems will not be displayed here, because I think barely watching these pictures delivers more feelings.
DeepL is used to translate my poems from Chinese to English. Stable Diffusion v1.5 and v2.1 are used to generate images based poems.




















