At the end of the year, domestic AI has really taken off. Today, Keling AI officially announced the launch of a multi-image reference feature that supports referencing multiple subjects simultaneously. Users can upload 1 to 4 reference images and select characters, objects, props, or scenes within them. By combining these references with text descriptions, the model will create imaginative videos, helping users achieve more personalized creations.
This feature is particularly suitable for short film production that requires visual consistency. Users can upload multiple images of the same character to ensure consistency across different shots. You can also specify scenes, costumes, actions, etc., through reference images, allowing characters to perform specific actions in particular environments.
For example, if you upload an image of a beautiful woman and then provide an image of a piece of jewelry and a stage scene, you can describe a scenario where she walks the runway wearing that jewelry.
Official creative description: On stage, a girl dressed in fashionable clothing, wearing a crystal crown, calmly looks at the camera.
Similarly, everyone can use this reference feature to have designated models showcase specific outfits and more.
Not only for characters, but you can also dress up your cat in cool outfits for a display:
Official creative description: An anthropomorphic cat wearing a jacket and sunglasses turns around on stage and poses for the camera.
This cartoon-style characters and scenes are also completely fine, making animation production much easier in the future:
Official creative description: A cartoon-style old man in a café lifts a cup to drink coffee.
