StudioAI is a fully functional Generative AI application built with Streamlit. This project demonstrates an end-to-end image generation pipeline utilizing the powerful Stable Diffusion v1.5 model.
Developed as a capstone project for the Microsoft ElevAte Training Center 2025, this application features both standard Text-to-Image generation and advanced Image-to-Image editing capabilities (Inpainting and seamless Outpainting/Zoom Out).
- Text-to-Image Generation: Generate high-quality images from descriptive text prompts.
- Advanced Schedulers: Support for multiple diffusion schedulers for fine-tuned generation.
- Interactive Inpainting: Draw masks directly on the web interface to edit, replace, or add specific objects to an existing image.
- Seamless Outpainting (Zoom Out): Expand the canvas of an image to reveal more of the surrounding environment using custom color-blocking and overlap masking techniques.
- Cloud-Ready: Clean architecture separated into logic and UI, fully optimized for deployment on Streamlit Community Cloud.
- Live App: [Insert your Streamlit App Link Here]
- Video Demo: Watch
video_demo_aplikasi_BFGAI.mp4in this repository for a full walkthrough.
- Frontend: Streamlit, Streamlit Drawable Canvas
- Machine Learning Framework: PyTorch
- Generative AI Models: Hugging Face
diffusers(StableDiffusionPipeline,StableDiffusionInpaintPipeline) - Image Processing: Pillow (PIL), NumPy
If you want to run this application locally on your machine, follow these steps:
git clone [https://github.com/yourusername/studio-ai-stable-diffusion.git](https://github.com/yourusername/studio-ai-stable-diffusion.git)
cd studio-ai-stable-diffusion
It is highly recommended to use a virtual environment.
pip install -r requirements.txt
streamlit run app.py
Note: A GPU (CUDA-enabled) is highly recommended for faster inference times.
📦 studio-ai-stable-diffusion
┣ 📜 app.py # Main Streamlit application interface
┣ 📜 logic.py # Core pipeline logic (Model loading, Txt2Img, Img2Img)
┣ 📜 requirements.txt # Python dependencies
┣ 📜 README.md # Project documentation
┗ 📓 Pipeline.ipynb # Step-by-step experimental notebook
- Live App: [click here!]
- Video Demo:
Erlangga Azhar * Tech Enthusiast
- Connect with me on LinkedIn
This project was developed as part of the Microsoft ElevAte Training Center 2025 curriculum.