|
--- |
|
pipeline_tag: object-detection |
|
tags: |
|
- ultralytics |
|
- yolo |
|
- yolov8 |
|
- tracking |
|
- image-classification |
|
- obb |
|
- object-detection |
|
language: |
|
- hy |
|
datasets: |
|
- armvectores/handwritten_text_detection |
|
--- |
|
# YOLOv8 Handwritten Text Detection |
|
|
|
|
|
## Model Description |
|
|
|
YOLOv8 is the eighth version of the You Only Look Once (YOLO) object detection algorithm. It excels in speed and accuracy, making it an ideal choice for real-time applications. The YOLOv8 model provided here has been fine-tuned on a diverse dataset of handwritten texts to improve its specificity in detecting handwritten content as opposed to typed or printed materials. |
|
|
|
|
|
## How to use |
|
|
|
``` |
|
pip install ultralytics |
|
``` |
|
|
|
``` |
|
from ultralytics import YOLO |
|
from huggingface_hub import hf_hub_download |
|
from matplotlib import pyplot as plt |
|
|
|
# Load the weights from our repository |
|
model_path = hf_hub_download(local_dir=".", |
|
repo_id="armvectores/yolov8n_handwritten_text_detection", |
|
filename="best.pt") |
|
model = YOLO(model_path) |
|
|
|
# Load test blank |
|
test_blank_path = hf_hub_download(local_dir=".", |
|
repo_id="armvectores/yolov8n_handwritten_text_detection", |
|
filename="test_blank.png") |
|
|
|
# Do the predictions |
|
res = model.predict(source=test_blank_path, project='.',name='detected', exist_ok=True, save=True, show=False, show_labels=False, show_conf=False, conf=0.5, ) |
|
|
|
plt.figure(figsize=(15,10)) |
|
plt.imshow(plt.imread('detected/test_blank.png')) |
|
plt.show() |
|
``` |
|
|
|
|
|
## Tests |
|
|
|
<p align="center"> |
|
<img width="400px" src="prediction1.png" alt="qr"/> |
|
</p> |
|
|
|
<p align="center"> |
|
<img width="400px" src="prediction2.png" alt="qr"/> |
|
</p> |
|
|
|
## Metrics |
|
|
|
The final IoU=0.98 |
|
|
|
The IoU during training |
|
<p align="center"> |
|
<img src="results.png" width="200" /> |
|
</p> |
|
|
|
|
