AI-SCANNER

• 

  3-D view with auto rotate ofcourse

• 

  wall -image i used ..try it for ur wall

Inspiration

Wanted to build a simple wall-scanner tool that uses AI to detect objects through images, making safety checks easier and more accessible.

What it does

Uses a YOLO model in the browser to analyze images and highlight potential objects behind walls using fast on-device inference.

How I built it

Built with React, ONNX Runtime Web, and a lightweight YOLO model, integrated into a clean UI with client-side processing.

Challenges I ran into

Handling model loading in the browser, optimizing performance, and fixing build issues with ONNX Runtime on deployment.

Accomplishments that I'm proud of

Running real-time AI inference entirely in the browser and achieving a working prototype with smooth user interaction.

What I learned

Working with ONNX models in web apps, deploying AI tools with React, and solving wasm-related build challenges.

What's next for AI-SCANNER

Adding live camera scanning, better accuracy models, and enhanced visualisation for safer and more reliable wall detection and training data, utilising multiple models for accurate classification.

Built With

• a
• and-a-lightweight-yolo-model
• clean
• client-side
• integrated
• into
• onnx-runtime-web
• react
• ui
• with