Hi, this is another attempt at AI&Alphonse Maria Mucha, the software used is still Midjourney, Procreat for iPad, and photoshop. Will the V4 model released by Midjourney make the whole image more logical? Will its generated style fits the input style better and expresses better story? I think this project answers the question somehow.
First, I input the prompt "A handsome man kneeling in the flower field, Alphonse Maria Mucha" into Midjourney's V4 model and get the following image:
The image of prompt "A handsome man kneeling in the flower field, Alphonse Maria Mucha" by Midjourney V4 model
It can be seen that, compared with the previous models, the completion level of the face and hand has been greatly improved in the preliminary generated image. Although the logic of showing the crossed fingers is not perfect, the "correct feeling" of the hand can be vaguely found in the result.
So what I did after was basically:
Fixed and completed lost, illogical, and wrong perspective parts.
Fixed "ambiguous" and random cluttered colors in the image.
Sharpened the lines to make the whole picture neat.
Expanded screen scale from 1:1 to 9:16.
Fixed and completed lost, illogical, and wrong perspective parts.
Fixed "ambiguous" and random cluttered colors in the image.
Sharpened the lines to make the whole picture neat.
Expanded screen scale from 1:1 to 9:16.
Adjusted the color tone to make the result more 'Mucha'.
And this is my final result :)
FInal Result
More results will be put in the future, you are so welcome to leave any comments&suggestions <3
