Struggling with LeonardoAI Image-to-Image in Cursor
Merry Christmas everyone! šāØ Iāve been trying all day to get the LeonardoAI API running inside my app using Cursor, but without success so far. My goal is to transform an uploaded image of a person into a completely different scene (classic image-to-image transformation), similar to what @Corbin Brown does with Thumio. In some cases, images do get generated, but the results are very poor: - the scene described in the prompt is basically not present at all, and - the personās face looks completely different from the original. - So essentially, it does the exact opposite of what itās supposed to do. Does anyone have experience with this and know what might cause these issues or what I should pay special attention to? Iām building the app using Cursorās AI Agent and Iām still relatively new to this field, so my technical knowledge is somewhat limited. Is LeonardoAI even the right API for this use case, or has anyone had better results with other image-to-image APIs? Thanks a lot in advance, and happy holidays! š
š