FusionBrain.AI is an open-source artificial intelligence model from abroad, known as the Kandinsky 2.1 model. The platform provides a free AI painting interactive portal, where users can generate images, expand paintings, blend images, and locally repaint images on a canvas.
FusionBrain's AI painting has its own unique features. Instead of generating individual images, it allows users to position and generate paintings on a large canvas using selection areas. Users can paint in different positions on the canvas, expand the painting in a specific area, and seamlessly blend different regions of the painting. Users can also upload other artworks and merge them with the current canvas content using AI.
I. Access and Interface
- Access the website: https://fusionbrain.ai/. The website does not require registration and can be used without any limitations. The website interface is shown in the image below. Simply click the "Open editor" button to enter the painting interface.
- Interface Introduction: The painting interface, as shown in the image below, consists of a canvas in the middle. There is a resizable and draggable selection area on top, which helps determine the painting position on the canvas. Below the canvas is a text prompt input box and a create button. The bottom left corner contains options for selecting, moving, and uploading photos, as well as smudging tools. The top right corner includes buttons for undoing, redoing, and downloading.
II. Creating and Appreciating Artworks
-
Oil Painting - Beach
Find a description of the scenery: an absolutely fascinating sunset on the beach, with a mix of orange, pink, and yellow in the sky. The clear sea gently kisses the coast, and the white sandy beach stretches endlessly. The scene is full of dynamism, with seagulls soaring high in the sky and palm trees swaying gently. Immerse yourself in a peaceful atmosphere. After translating the description into English, input the text prompt and create the artwork. The result is shown in the image below.
-
Oil Painting - Mountains and Rivers
Adjust the selection area and input the text prompt: "poetic scenery far and wide, bright sunlight, with twisting rivers and karst landscape, perspective, drone view, 8k, beautiful sunrise, --ar 9:5." After submitting, the artwork is generated as shown in the image below.
- Extending the Painting: Expand the selection area and repaint the blank area outside the image. The generated content is shown in the image below.
-
Blending Paintings - Christmas Style
Overlap the selection area with an existing painting. Input the text prompt: "Mountains, pine trees, white snow." After creating, the artwork is shown in the image below.
- Image-to-Image Generation: Upload a photo and smudge the background. After creating, the artwork is shown in the image below.
Note:
-
There are already many platforms for AI-generated images, but generating multiple images on a single canvas is rare, and it can be extended to various applications such as image blending, expansion, and repainting.
-
The platform offers 23 different painting art styles, providing a rich variety of artwork types.