Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,32 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
tags:
|
| 4 |
+
- object-detection
|
| 5 |
+
- vision
|
| 6 |
+
- yolov8
|
| 7 |
+
- sports
|
| 8 |
+
- football
|
| 9 |
+
- tennis
|
| 10 |
+
library_name: ultralytics
|
| 11 |
+
pipeline_tag: object-detection
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
# ⚽ Football & Sports Player Detection (YOLOv8)
|
| 15 |
+
|
| 16 |
+
This model is a fine-tuned version of **YOLOv8 Nano**, designed to detect players, balls, and sports equipment in match footage (Football/Tennis).
|
| 17 |
+
|
| 18 |
+
It was developed specifically to demonstrate **Data Engineering** techniques, achieving high generalization performance even when trained on a **restricted dataset** through the use of aggressive offline and online Data Augmentation strategies.
|
| 19 |
+
|
| 20 |
+
## 🏷️ Classes
|
| 21 |
+
The model is trained to detect the following 3 classes:
|
| 22 |
+
* `0`: **Person** (Players, Referees)
|
| 23 |
+
* `1`: **Ball** (Football, Tennis ball)
|
| 24 |
+
* `2`: **Equipment** (Rackets, Goal posts, etc.)
|
| 25 |
+
|
| 26 |
+
## 💻 How to Use
|
| 27 |
+
|
| 28 |
+
To use this model in your Python project, you need the `ultralytics` and `huggingface_hub` libraries.
|
| 29 |
+
|
| 30 |
+
### Installation
|
| 31 |
+
```bash
|
| 32 |
+
pip install ultralytics huggingface_hub
|