| --- |
| license: apache-2.0 |
| datasets: |
| - HyzeAI/HyzeACR-Dataset |
| language: |
| - en |
| pipeline_tag: image-classification |
| tags: |
| - HyzeAI |
| - HyzeACR |
| - HiteshV |
| - MadeByKids |
| - OpenSource |
| - Space |
| - Astronomy |
| --- |
| |
| <p align="center"> |
| <img src="https://i.ibb.co/99ybtt8Y/Hyze-ACR-Banner.png" alt="HyzeACR" width="650"/> |
| </p> |
|
|
| <h1 align="center">HyzeACR</h1> |
|
|
| <p align="center"> |
| A lightweight image-classification model by <b>HyzeAI</b> |
| </p> |
|
|
| <p align="center"> |
| <a href="https://hyze.dev">Chat with all models</a> • |
| <a href="https://hyzeacr.netlify.app">HyzeACR (Web Demo)</a> |
| </p> |
|
|
| --- |
|
|
| ## The Live Demo |
| Go to [https://hyzeacr.netlify.app](https://hyzeacr.netlify.app) |
| --- |
|
|
| ## What This Project Does |
|
|
| This AI system classifies space related images into the following categories: |
|
|
| - Moons |
| - Planets |
| - Galaxies |
| - Nebulae |
|
|
| It supports: |
| - Image based classification |
| --- |
|
|
| ## How It Works |
|
|
| 1. A trained machine learning model is loaded in the browser using TensorFlow.js |
| 2. Any image of a Moon, Planet, Nebulae, or a Galaxy is uploaded to the model |
| 3. The model predicts the most likely space object |
| 4. The predictions are displayed |
|
|
| --- |
|
|
| ## How to use |
|
|
| 1. The easiest way to use the model is by using the web demo at [https://hyzeacr.netlify.app](https://hyzeacr.netlify.app) (The model is hosted with Google Cloud) |
| 2. Install a local TensorFlow/Keras environment |
| 3. Run this command pip install tensorflow numpy |
| 4. Next write a python script to run the model |
|
|
| --- |
|
|
| ## Tech Stack |
|
|
| - TensorFlow.js (browser inference) |
| - Keras (model training) |
| - TensorFlow (ML framework) |
| - JavaScript (frontend logic) |
| - HTML (UI) |
|
|
| --- |
|
|
| ## Created by Hitesh Vinothkumar |
