Beanbag then generated me 10 images of the same character in 10 different scenes
Previously, shared i.e. Dream 3.0 Keeping the CharacterconsistencyThe hands-on approach. Pages that need it can check it out by clicking on the link below.
How to write character consistency cue words? Instant Dream AI Video Generation to Keep Character Consistency
Today's Sharing
- A new take on beanbag ai character consistency
single character
Based on the previous post, the individual character charts of our heroes and heroines, specifically theVincent FigureFor descriptors, check out the previous post
Then we open the beanbag ai: https://www.doubao.com/
Upload a picture and enter a cue word:Generate 10 images of different actions and scenes around the character, keeping the character's role and style and dress uniform and consistent
Beanbag then generated me 10 images of the same character in 10 different scenes
Isn't it amazing how basically consistent both the action, the characters, and the costumes are!!!!
And look at our heroine.
There's a bit of a discrepancy in the heroine's costume, more or less. I hope Doubao can continue to improve ！！！！
dual role (in sports)
Tried the dual-subject birth chart out of curiosity
Prompt words:Based on the above two personas, generate 10 interactive scenes of them in the same frame
666, the effect is absolute!
Not only is the image style unified, but also a key to handle the two people in the same frame interactive scenes, but also to generate video.
at last
Methods of maintaining character consistency in i.e. Dream AI: through [Character Seeds] and [Reference Chart
Beanbag AI can be used to access different scenes and actions directly through the graph-born method, and it can also be applied to dual subjects! This is very friendly to the little white man!