celestial-mini / README.md
langutang's picture
Update README.md
cd81677 verified
---
license: mit
language:
- en
library_name: transformers
tags:
- cv
- robotics
pipeline_tag: object-detection
---
# 🌌 Celestial-Mini: Lightweight Object Detection Model (TF)
[![TensorFlow](https://img.shields.io/badge/framework-TensorFlow-orange)](https://www.tensorflow.org/)
[![Object Detection](https://img.shields.io/badge/task-Object%20Detection-blue)]()
[![Models](https://img.shields.io/badge/targets-80%20Objects-green)]()
**Celestial-Mini** is a compact, high-performance object detection model designed to recognize up to **80 distinct object classes**. Built with **TensorFlow**, it balances speed and accuracy for deployment in edge devices and real-time applications.
---
## πŸš€ Key Features
- πŸ” Detects up to **80 different object categories**
- ⚑ Optimized for **real-time inference**
- 🧠 Built on a **lightweight backbone**
- πŸ“¦ TensorFlow SavedModel format for easy deployment
- 🧰 Compatible with TensorFlow Lite and TensorFlow.js
---
## πŸ§ͺ Intended Use
Celestial-Mini is designed for:
- Robotics and drones
- Smart home devices
- Augmented Reality (AR) systems
- Mobile applications
- Educational and prototyping environments
---
## 🏷 Object Classes
Includes detection support for the standard 80-class COCO-style object categories such as:
```
person, bicycle, car, motorcycle, airplane, bus, train, truck, boat, traffic light, ...
```
---
## πŸ“¦ How to Use
```python
import tensorflow as tf
# Load the model
model = tf.saved_model.load("path/to/celestial-mini")
# Run inference
detections = model(input_tensor)
```
---
## πŸ“Š Performance
| Metric | Value |
|----------------|---------------|
| Classes | 80 |
| Model Size | ~15MB |
| Inference Time | < 50ms/image |
| Framework | TensorFlow |
> πŸ“Œ Performance may vary depending on hardware and TensorFlow backend optimizations.
---
## 🧠 Training & Dataset
Celestial-Mini was trained on a custom variant of the **COCO dataset**, emphasizing generalization and real-time inference. Model architecture includes quantization-friendly layers and depthwise separable convolutions.
---
## πŸ“– Citation
If you use **Celestial-Mini** in your work, please consider citing:
```
@misc{celestialmini2025,
title={Celestial-Mini: A Lightweight Real-Time Object Detector},
author={Lang, John},
year={2025},
howpublished={\url{https://huggingface.co/langutang/celestial-mini}}
}
```
---
## πŸ“¬ Contact & License
- πŸ“« For questions or collaboration, open an issue or contact the maintainer.
- βš–οΈ License: MIT (see LICENSE file for details)
---
## 🌠 Hugging Face Model Hub
To load from Hugging Face:
```python
from transformers import AutoFeatureExtractor, TFModelForObjectDetection
model = TFModelForObjectDetection.from_pretrained("langutang/celestial-mini")
extractor = AutoFeatureExtractor.from_pretrained("langutang/celestial-mini")
```
---
Transform your edge AI projects with the power of **Celestial-Mini** πŸŒ