# YOLOv8 Detection

Run real-time object detection with Ultralytics YOLOv8 and YOLOv11.

{% hint style="success" %}
All examples can be run on GPU servers rented through [CLORE.AI Marketplace](https://clore.ai/marketplace).
{% endhint %}

{% hint style="info" %}
**Update: YOLOv11 (2025) — 22% Faster**

YOLOv11 is now available via the same `ultralytics` package. It delivers **22% faster inference** and improved mAP over YOLOv8, with the same simple API. New features include Oriented Bounding Box (OBB) detection. Upgrade by running `pip install -U ultralytics`.
{% endhint %}

## Renting on CLORE.AI

1. Visit [CLORE.AI Marketplace](https://clore.ai/marketplace)
2. Filter by GPU type, VRAM, and price
3. Choose **On-Demand** (fixed rate) or **Spot** (bid price)
4. Configure your order:
   * Select Docker image
   * Set ports (TCP for SSH, HTTP for web UIs)
   * Add environment variables if needed
   * Enter startup command
5. Select payment: **CLORE**, **BTC**, or **USDT/USDC**
6. Create order and wait for deployment

### Access Your Server

* Find connection details in **My Orders**
* Web interfaces: Use the HTTP port URL
* SSH: `ssh -p <port> root@<proxy-address>`

## What is YOLOv8?

YOLOv8 is a high-performance YOLO model offering:

* Object detection
* Instance segmentation
* Pose estimation
* Image classification
* Object tracking

## What is YOLOv11?

YOLOv11 (2025) is the latest generation, adding:

* **22% faster inference** vs YOLOv8
* Higher mAP across all model sizes
* **Oriented Bounding Box (OBB)** detection — new task
* Improved architecture (C3k2 blocks, SPPF, C2PSA)
* Same `ultralytics` package, drop-in replacement

### Supported Tasks (YOLOv11)

| Task       | Suffix   | Description                                         |
| ---------- | -------- | --------------------------------------------------- |
| `detect`   | *(none)* | Object detection with bounding boxes                |
| `segment`  | `-seg`   | Instance segmentation with masks                    |
| `classify` | `-cls`   | Image classification                                |
| `pose`     | `-pose`  | Human pose estimation                               |
| `obb`      | `-obb`   | **NEW** Oriented bounding boxes (rotated detection) |

## Model Sizes

### YOLOv8 Models

| Model   | Size  | mAP  | Speed (RTX 3090) |
| ------- | ----- | ---- | ---------------- |
| YOLOv8n | 3.2M  | 37.3 | \~1ms            |
| YOLOv8s | 11.2M | 44.9 | \~2ms            |
| YOLOv8m | 25.9M | 50.2 | \~4ms            |
| YOLOv8l | 43.7M | 52.9 | \~6ms            |
| YOLOv8x | 68.2M | 53.9 | \~8ms            |

### YOLOv11 Models

| Model   | Size  | mAP  | Speed (RTX 3090) |
| ------- | ----- | ---- | ---------------- |
| yolo11n | 2.6M  | 39.5 | \~0.8ms          |
| yolo11s | 9.4M  | 47.0 | \~1.5ms          |
| yolo11m | 20.1M | 51.5 | \~3.2ms          |
| yolo11l | 25.3M | 53.4 | \~4.7ms          |
| yolo11x | 56.9M | 54.7 | \~6.5ms          |

### YOLOv8 vs YOLOv11 Comparison

| Metric                | YOLOv8x | yolo11x | Improvement      |
| --------------------- | ------- | ------- | ---------------- |
| Parameters            | 68.2M   | 56.9M   | **-17% smaller** |
| mAP50-95 (COCO)       | 53.9    | 54.7    | **+0.8 mAP**     |
| Inference (RTX 3090)  | \~8ms   | \~6.5ms | **+22% faster**  |
| FPS (RTX 3090, 640px) | \~150   | \~183   | **+22% faster**  |
| OBB Task              | ❌       | ✅       | **New in v11**   |

## Quick Deploy

**Docker Image:**

```
pytorch/pytorch:2.5.1-cuda12.4-cudnn9-runtime
```

**Ports:**

```
22/tcp
7860/http
```

**Command (YOLOv11):**

```bash
pip install ultralytics gradio && \
python -c "
import gradio as gr
from ultralytics import YOLO
from PIL import Image

model = YOLO('yolo11m.pt')

def detect(image):
    results = model(image)
    return Image.fromarray(results[0].plot())

demo = gr.Interface(fn=detect, inputs=gr.Image(type='pil'), outputs=gr.Image(), title='YOLOv11 Detection')
demo.launch(server_name='0.0.0.0', server_port=7860)
"
```

## Accessing Your Service

After deployment, find your `http_pub` URL in **My Orders**:

1. Go to **My Orders** page
2. Click on your order
3. Find the `http_pub` URL (e.g., `abc123.clorecloud.net`)

Use `https://YOUR_HTTP_PUB_URL` instead of `localhost` in examples below.

## Installation

```bash
pip install ultralytics
```

Same package for both YOLOv8 and YOLOv11. Upgrade to get YOLOv11:

```bash
pip install -U ultralytics
```

## YOLOv11 Object Detection

### Basic Detection with yolo11m

```python
from ultralytics import YOLO
from PIL import Image

# Load YOLOv11 medium model
model = YOLO('yolo11m.pt')

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

# Show results
results[0].show()

# Save results
results[0].save('output.jpg')
```

### Get Detections

```python
results = model('image.jpg')

for result in results:
    boxes = result.boxes
    for box in boxes:
        # Coordinates
        x1, y1, x2, y2 = box.xyxy[0].tolist()

        # Confidence
        conf = box.conf[0].item()

        # Class
        cls = int(box.cls[0].item())
        name = model.names[cls]

        print(f"{name}: {conf:.2f} at ({x1:.0f}, {y1:.0f}, {x2:.0f}, {y2:.0f})")
```

### Batch Processing

```python
from ultralytics import YOLO
import os

model = YOLO('yolo11m.pt')

input_dir = './images'
output_dir = './detected'
os.makedirs(output_dir, exist_ok=True)

# Process all images
results = model(input_dir, save=True, project=output_dir)
```

## YOLOv11 Tasks

### Instance Segmentation

```python
from ultralytics import YOLO

# Load YOLOv11 segmentation model
model = YOLO('yolo11m-seg.pt')

results = model('image.jpg')

for result in results:
    masks = result.masks  # Segmentation masks
    if masks is not None:
        for mask in masks.data:
            # mask is a binary tensor
            pass
```

### Pose Estimation

```python
from ultralytics import YOLO

# Load YOLOv11 pose model
model = YOLO('yolo11m-pose.pt')

results = model('image.jpg')

for result in results:
    keypoints = result.keypoints
    if keypoints is not None:
        for kp in keypoints.data:
            # 17 keypoints: nose, eyes, ears, shoulders, elbows, wrists, hips, knees, ankles
            pass
```

### Classification

```python
from ultralytics import YOLO

# Load YOLOv11 classification model
model = YOLO('yolo11m-cls.pt')

results = model('image.jpg')
for result in results:
    # Top-1 class and confidence
    top1 = result.probs.top1
    top1conf = result.probs.top1conf.item()
    print(f"Class: {result.names[top1]} ({top1conf:.2f})")
```

### Oriented Bounding Box (OBB) — NEW in YOLOv11

OBB detects objects at any rotation angle — perfect for aerial/satellite imagery, document scanning, and text detection.

```python
from ultralytics import YOLO

# Load YOLOv11 OBB model
model = YOLO('yolo11m-obb.pt')

results = model('aerial_image.jpg')

for result in results:
    obb = result.obb
    if obb is not None:
        for box in obb:
            # Rotated box: x_center, y_center, width, height, angle
            xywhr = box.xywhr[0].tolist()
            conf = box.conf[0].item()
            cls = int(box.cls[0].item())
            print(f"{result.names[cls]}: {conf:.2f} rotated box: {xywhr}")
```

## Video Processing

### Process Video

```python
from ultralytics import YOLO

model = YOLO('yolo11m.pt')

# Process video
results = model('video.mp4', save=True)
```

### Real-Time Webcam

```python
from ultralytics import YOLO
import cv2

model = YOLO('yolo11n.pt')  # Use nano model for real-time

cap = cv2.VideoCapture(0)

while True:
    ret, frame = cap.read()
    if not ret:
        break

    results = model(frame)
    annotated = results[0].plot()

    cv2.imshow('YOLOv11', annotated)
    if cv2.waitKey(1) & 0xFF == ord('q'):
        break

cap.release()
cv2.destroyAllWindows()
```

### Save Processed Video

```python
from ultralytics import YOLO
import cv2

model = YOLO('yolo11m.pt')

cap = cv2.VideoCapture('input.mp4')
fps = cap.get(cv2.CAP_PROP_FPS)
width = int(cap.get(cv2.CAP_PROP_FRAME_WIDTH))
height = int(cap.get(cv2.CAP_PROP_FRAME_HEIGHT))

out = cv2.VideoWriter('output.mp4', cv2.VideoWriter_fourcc(*'mp4v'), fps, (width, height))

while cap.isOpened():
    ret, frame = cap.read()
    if not ret:
        break

    results = model(frame)
    annotated = results[0].plot()
    out.write(annotated)

cap.release()
out.release()
```

## Object Tracking

```python
from ultralytics import YOLO

model = YOLO('yolo11m.pt')

# Track objects in video
results = model.track('video.mp4', save=True, tracker='bytetrack.yaml')

# Access tracking IDs
for result in results:
    boxes = result.boxes
    if boxes.id is not None:
        track_ids = boxes.id.tolist()
```

## Custom Training

### Prepare Dataset

```yaml

# dataset.yaml
path: /workspace/dataset
train: images/train
val: images/val

names:
  0: cat
  1: dog
  2: bird
```

### Train YOLOv11

```python
from ultralytics import YOLO

# Load pretrained YOLOv11 model
model = YOLO('yolo11n.pt')

# Train
results = model.train(
    data='dataset.yaml',
    epochs=100,
    imgsz=640,
    batch=16,
    device=0
)
```

### Training Arguments

```python
model.train(
    data='dataset.yaml',
    epochs=100,
    imgsz=640,
    batch=16,
    device=0,
    workers=8,
    patience=50,         # Early stopping
    save=True,
    save_period=10,      # Save every N epochs
    cache=True,          # Cache images
    amp=True,            # Mixed precision
    lr0=0.01,
    lrf=0.01,
    momentum=0.937,
    weight_decay=0.0005,
    warmup_epochs=3.0,
    box=7.5,
    cls=0.5,
    dfl=1.5,
    augment=True,
    hsv_h=0.015,
    hsv_s=0.7,
    hsv_v=0.4,
    flipud=0.0,
    fliplr=0.5,
    mosaic=1.0,
    mixup=0.0,
)
```

## Export Model

```python
from ultralytics import YOLO

model = YOLO('yolo11m.pt')

# Export to different formats
model.export(format='onnx')           # ONNX
model.export(format='tensorrt')       # TensorRT
model.export(format='openvino')       # OpenVINO
model.export(format='coreml')         # CoreML
model.export(format='tflite')         # TensorFlow Lite
```

## API Server

```python
from fastapi import FastAPI, UploadFile
from fastapi.responses import JSONResponse
from ultralytics import YOLO
from PIL import Image
import io

app = FastAPI()
model = YOLO('yolo11m.pt')

@app.post("/detect")
async def detect(file: UploadFile):
    contents = await file.read()
    image = Image.open(io.BytesIO(contents))

    results = model(image)

    detections = []
    for box in results[0].boxes:
        detections.append({
            "class": model.names[int(box.cls[0])],
            "confidence": float(box.conf[0]),
            "bbox": box.xyxy[0].tolist()
        })

    return JSONResponse(content={"detections": detections})

@app.post("/segment")
async def segment(file: UploadFile):
    contents = await file.read()
    image = Image.open(io.BytesIO(contents))

    model_seg = YOLO('yolo11m-seg.pt')
    results = model_seg(image)

    # Return annotated image
    annotated = results[0].plot()
    # ... convert and return

# Run: uvicorn server:app --host 0.0.0.0 --port 8000
```

## Performance Optimization

### TensorRT Export

```python
model = YOLO('yolo11m.pt')
model.export(format='tensorrt', half=True)

# Use exported model
model_trt = YOLO('yolo11m.engine')
results = model_trt('image.jpg')
```

### Batch Inference

```python

# Process multiple images at once
images = ['img1.jpg', 'img2.jpg', 'img3.jpg', 'img4.jpg']
results = model(images, batch=4)
```

## Performance Benchmarks

### YOLOv11 FPS (640px input)

| Model   | GPU      | FPS    |
| ------- | -------- | ------ |
| yolo11n | RTX 3090 | \~1100 |
| yolo11s | RTX 3090 | \~730  |
| yolo11m | RTX 3090 | \~370  |
| yolo11x | RTX 3090 | \~183  |
| yolo11x | RTX 4090 | \~305  |

### YOLOv8 FPS (640px input) — Previous Generation

| Model   | GPU      | FPS   |
| ------- | -------- | ----- |
| YOLOv8n | RTX 3090 | \~900 |
| YOLOv8s | RTX 3090 | \~600 |
| YOLOv8m | RTX 3090 | \~300 |
| YOLOv8x | RTX 3090 | \~150 |
| YOLOv8x | RTX 4090 | \~250 |

## Troubleshooting

### Out of Memory

```python

# Use smaller model
model = YOLO('yolo11n.pt')

# Or reduce image size
results = model('image.jpg', imgsz=320)
```

### Slow Processing

* Use TensorRT export
* Use smaller model (yolo11n or yolo11s)
* Reduce image size

### Low Accuracy

* Use larger model (yolo11x instead of yolo11n)
* Train on custom data
* Increase image size

## Cost Estimate

Typical CLORE.AI marketplace rates (as of 2025):

| GPU       | Hourly Rate | Daily Rate | 4-Hour Session |
| --------- | ----------- | ---------- | -------------- |
| RTX 3060  | \~$0.03     | \~$0.70    | \~$0.12        |
| RTX 3090  | \~$0.06     | \~$1.50    | \~$0.25        |
| RTX 4090  | \~$0.10     | \~$2.30    | \~$0.40        |
| A100 40GB | \~$0.17     | \~$4.00    | \~$0.70        |
| A100 80GB | \~$0.25     | \~$6.00    | \~$1.00        |

*Prices vary by provider and demand. Check* [*CLORE.AI Marketplace*](https://clore.ai/marketplace) *for current rates.*

**Save money:**

* Use **Spot** market for flexible workloads (often 30-50% cheaper)
* Pay with **CLORE** tokens
* Compare prices across different providers

## Next Steps

* [Segment Anything](https://docs.clore.ai/guides/image-processing/segment-anything) - Advanced segmentation
* Detectron2 - More detection options
* [Real-ESRGAN](https://docs.clore.ai/guides/image-processing/real-esrgan-upscaling) - Enhance detected objects