Interior-stable-difusion: Revolutionizing Interior Design with Rapid Visualization and Customization

The evolution of AI technologies like Stable Diffusion(https://arxiv.org/abs/2112.10752) has revolutionized visual design. Now, with "Interior-Stable-Diffusion," this technology is tailored for interior design, enabling rapid generation, style modification, and object replacement in interior spaces. This application empowers designers to visualize and refine spaces with unprecedented speed and precision, transforming ideas into reality in moments.

Main Functions

Using Stable Diffusion, this app can make desirable images with three main function:

• General generation: Utilize the text2img pipeline to create detailed interior images from textual descriptions.
• Fixing style: Utilize the Controlnet Canny echnique to maintain the original image’s edges while introducing a new style based on the canny edge map.
• Replacing object: using ControlnetInpaintPipeline to seamlessly replace objects within specified masked areas of an image.

Examples

General generation

Prompt: A living room with a TV, wooden floor, a sofa, a nice glass table and a flower in the table

Prompt: A large modern kitchen with light grey, brown and white, large kitchen cabinets

Fixing style

Change: A black table

Change: A colorful violet chandelier, darker ceiling.

Replacing object

Prompt: a fridge

Prompt: a luxury liquor cabinet

Installation and Usage

Environment setup

python3 -m venv .env
source .env/bin/activate

git clone https://github.com/Trgtuan10/Interior-stable-difusion.git
cd Interior-stable-difusion
pip install -r requirements.txt

Download my checkpoint

mkdir checkpoints
cd checkpoints
wget https://civitai.com/api/download/models/128713 -O Interior.safetensors
wget https://civitai.com/api/download/models/195419 -O Interior_lora.safetensors

Run app

cd App_demo
streamlit run streamlit_app