Skip to content

Object Scope — Browser-Based Real-Time Object Detection with TensorFlow.js #8745

Description

@fouad-salehi

Hello TensorFlow.js community,

I would like to share Object Scope, a browser-based computer vision project built with TensorFlow.js and COCO-SSD.

Live Demo: https://objectscope-research.web.app/
Source Code: https://github.com/fouad-salehi/objectScope

What is Object Scope?

Object Scope demonstrates real-time and image-based object detection directly in the web browser, without requiring a dedicated backend for inference.

It currently provides two main modes:

Image Detection — Upload an image and detect objects.
Camera Detection — Use a live camera stream for real-time object detection.

The application displays detected objects with bounding boxes and confidence scores.

Technology
TensorFlow.js
COCO-SSD
JavaScript
HTML5 / CSS3
Firebase Hosting
Why I'm sharing it

The goal of this project is to explore the practical use of TensorFlow.js for computer vision applications running entirely in the browser.

I would appreciate feedback from the TensorFlow.js community regarding:

Real-time detection performance
Browser compatibility
Model loading and inference performance
Potential improvements to the TensorFlow.js implementation
Suggestions for future development

If you have a few minutes, you can try the live demo directly in your browser and share your feedback.

Thank you to the TensorFlow.js team and community for the tools that made this project possible.

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

Type

No type

Projects

No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions