I Built an AI that Recreates Reality and it's INSANE!

In this video, I'll be showcasing my latest AI project where I built an AI that recreates reality, and the results are insane. Using a combination of object removal ai and object replacement ai techniques, I've created an ai image editing tool that can generate realistic images based on prompt based image editing. The ai photo editor utilizes stable diffusion inpainting and generative fill alternative methods to produce stunning results. I've also incorporated the sam model, mobilesam, and simplelama models into my project, which is built using openvino, fastapi, react, and vite. To handle the backend, I've used celery and redis, and deployed the model using hugging face spaces and vercel. This ai project demonstrates the power of machine learning and computer vision, and I'll be sharing my experience of building it in public, including the coding project and software engineering aspects. If you're interested in ai image generators, ai editors, or want to learn more about chatgpt tutorial and zeyadcode, then this video is for you. I'll be providing a comprehensive overview of my project, including the vandalizer tool, and how I used it to create realistic images. So, watch until the end to see the mind-blowing results of my ai project and learn more about the technology behind it.
I built a reality-editing AI image editor that can detect objects in photos, create precise masks, remove unwanted objects, blur them, or replace them with AI-generated content. In this video, I break down how I built a prompt-based image editing app using OWLv2 object detection, SAM / MobileSAM segmentation, and Stable Diffusion inpainting.

This is a full walkthrough of how text prompts can be turned into visual edits: OWLv2 finds the target object, SAM converts the bounding box into a mask, and the diffusion model reconstructs the image to remove or replace the selected area. I also show the real development process, including the cursed AI failures, debugging, deployment issues, and the final working demo.

The app is built as a full-stack AI project using FastAPI, React, Vite, Celery, Redis, OpenVINO, Hugging Face Spaces, and Vercel.

In this video, you will learn:

How to build an AI image editing app
How OWLv2 object detection works with text prompts
How SAM / MobileSAM segmentation creates accurate masks
How to use Stable Diffusion inpainting for object removal and replacement
How to build and deploy a full-stack AI app
How I handled the hardest parts of making the system actually work

If you are interested in computer vision, generative AI, image editing, object removal, AI inpainting, segmentation, or full-stack AI projects, this video will show you the complete pipeline from concept to deployment.

Tech stack: FastAPI, React, Vite, Celery, Redis, OpenVINO, Hugging Face Spaces, Vercel, OWLv2, SAM, MobileSAM, Stable Diffusion.
Website URL: https://vandalizer-frontend.vercel.app/
Github Repo: https://github.com/zeyad-shaban/vandalizer-frontend

Видео I Built an AI that Recreates Reality and it's INSANE! канала ZeyadCode