AI Speed Box

AI News and Information Link Collection (Mintbear's ignorant scraps, matured to become Visual AI News)
All
AI Image
AI Video
AI Sound
AI LLM
AI 3D
AI
AR, XR, VR
AI Toons
AI SNS

Kling Elements! (Multiple characters, appearing at the same time!)

Category
  1. AI Video
Gen
  1. Kling
Date
2025/01/21
Summary 🍀🧸
Kling Elements! 🍀🧸

Allows users to combine up to four images to create a video with a consistent character, environment, or object. (This feature is currently available for early access to select creative partners.)

Users can combine multiple static images to create visually appealing animated sequences.
URL
https://klingai.com
URL
https://klingai.com/release-notes
Release
Limited Release (Partial Release)
Informer
Report: Picassong

Kling Elements! 🍀🧸

Allows users to combine up to four images to create a video with a consistent character, environment, or object. (This feature is currently available for early access to select creative partners.)
Users can combine multiple static images to create visually appealing animated sequences.

🍀
Elements, this is a pretty important feature, because-

Even with current image generation technology, it is not easy to use multiple elements or multiple character references in a single image. Most of the time, one character reference is used, and most of the time, background images or styles are added to a single character.

However, when multiple elements in an AI video are consistent and multiple characters are directed consistently , the way AI films are produced changes completely.

Until now, only one character has appeared on one screen, or when two people appear, one has to turn around, or a directing method such as an over-the-shoulder shot has been used. It was not easy to consistently show two characters on one screen.

With this feature, it is now revolutionary and convenient. Cling, it's cool!

(As of midnight 2025.01.21, it has been delayed and will take approximately 90 minutes to generate. Some test results are at the bottom.)

Now you can use the Elements feature with the KLING AI 1.6 model to generate videos from images! Upload 1-4 images, select the subjects of the images (people, animals, objects or scenes) as elements, and describe their actions and interactions. A video will be generated based on the elements and prompts.

Here are some use cases for this feature:

1. Elements for subject consistency

Upload one or more images of your subject (people, characters, animals, objects, etc.) and we'll generate a video based on those image references with a consistent style. This can be especially useful for creating a consistent look for your character across different scenes.

You can also set up scenes or clothing as elements and assign actions to them in prompts to make your subject look and move in a certain way in a certain scene. This gives you more creative control over your work.

Sample 1: People + Objects + Space

Prompt: On the stage, a girl wearing fashionable clothes and a crystal crown calmly gazes at the camera.

Sample 2: Animal + Clothing + Accessories

Prompt: A standing cat character wearing a jacket and sunglasses strikes a pose towards the camera on the stage.

Sample 3: Animal + Clothing

Prompt: A white Bichon Frisé wearing a red floral Chinese-style winter coat licks its paw.
A white Bichon Frise wearing a red floral Chinese winter coat licks its front paw.

Sample 4: Anime Character + Space + Props

Prompt: In a café, a cartoon-style elderly man lifts a cup to drink coffee.
A cartoon-style old man in a cafe lifts a cup to drink coffee.

2. Character Interaction Elements

Upload multiple subjects (people, animals, objects) and describe their interactions.

Sample 5: People + People

Prompt: Two girls hug each other. Two girls hug each other.
It's a slightly different background and style, but it works. However, it's a bit awkward and has a dull tone and poor image quality :)

Sample 6: People + Animals

Prompt: A boy rides a Pegasus, soaring through the air, in a magical style.

Sample 7: Anime human character + anime animal character

Prompt: A cartoon Character wearing a white hat, and a cartoon-style bear, sitting side by side, wave and nudge at each other.
A cartoon character wearing a white hat and a cartoon-style bear sit side by side, waving at each other and playfully bumping elbows.

Test- waiting
For testing, I deliberately used four characters with completely different drawing styles.
Two of them are somewhat fleshed out, another one has movements, and another one is gone. But it's all very well done.
Two characters with different styles are also well directed. You can see the traces of efforts to match the color tones of the original image with the scene where the two characters meet. However, as expected from Kling, there is a slight toning down phenomenon. Still, this is great. Also, the directing effort to make the most of the attached image as a starting point is admirable. First of all, pass! Now let's build up the know-how!

Example of use

2
👍
1
😍
1
J
Jae Young Chang
음, 진짜 좋아지겠네요. 앞으로 더 많이 발전할 듯 합니다. 근데 중국산이라고 들었는데 중국산이 다른 것들보다 이렇게 좋다니....이건 인정해야 할 것 같아요
😀
1
👍
1
mintbear
네, 정말 그런 것 같아요. 거대한 대륙의 데이터로 학습된 Kling 과 Hailuo모델은 Sora 조차 따라가지 못하는 영역이 있는 듯 합니다.
See latest comments
Made with Slashpage