YOLOv5 for Layout Detection
Dataset
Dataset available in kaggle
Supported Labels
["caption", "chart", "image", "image_caption", "table", "table_caption", "text", "title"]
How to use
- Install library
pip install yolov5==7.0.5 torch
Load model and perform prediction
import yolov5
from PIL import Image
model = yolov5.load(models_id)
model.overrides['conf'] = 0.25 # NMS confidence threshold
model.overrides['iou'] = 0.45 # NMS IoU threshold
model.overrides['max_det'] = 1000 # maximum number of detections per image
# set image
image = 'https://huggingface.co./spaces/hermanshid/yolo-layout-detector-space/raw/main/test_images/example1.jpg'
# perform inference
results = model.predict(image)
# observe results
print(results[0].boxes)
render = render_result(model=model, image=image, result=results[0])
render.show()
- Downloads last month
- 37
Space using hermanshid/yolo-layout-detector 1
Evaluation results
- [email protected](box)self-reported0.979