Skip to main content

📦 Object Detection (YOLO)

Classification tells you "There is a Dog in this image." Object Detection tells you "There is a Dog, and it is located inside this exact Box [X, Y, Width, Height]."

🚀 You Only Look Once (YOLO)​

Older algorithms used to scan the image 2,000 times to find boxes. It was agonizingly slow. YOLO split the image into a grid and predicted all the boxes in a single, massive forward pass! It was so fast it could process real-time webcam video at 60 FPS!

🐍 Python Implementation​

Today, developers use the ultralytics library to run YOLO in 3 lines of code!

# pip install ultralytics
# from ultralytics import YOLO

# Load a pre-trained YOLOv8 model
# model = YOLO('yolov8n.pt')

# Run inference on an image
# results = model('my_dog_photo.jpg')

# Show the image with bounding boxes drawn!
# results[0].show()