Stock Market Pattern Detection YOLOv8s
foduucom/stockmarket-pattern-detection-yolov8
published Aug 2023 · updated Apr 2025
Stock Market Pattern Detection YOLOv8s is a detection model that identifies chart patterns in real-time from screen-captured stock market data.
specs
| Task | Object Detection (Stock Market Pattern Detection) |
| Architecture | YOLOv8s (You Only Look Once small variant) |
| Supported Labels | Head and shoulders bottom, Head and shoulders top, M_Head, StockLine, Triangle, W_Bottom |
| Input | Image (screen capture of stock market charts) |
| Output | Bounding boxes with predicted pattern labels |
about this model
YOLOv8s Stock Market Pattern Detection is an object detection model, based on the YOLOv8 framework, that identifies chart patterns in real time from screen-captured stock market charts.
Capabilities
The model detects six chart pattern classes:
- Head and shoulders bottom
- Head and shoulders top
- M_Head
- StockLine
- Triangle
- W_Bottom
It is designed to process live screen captures of trading charts, enabling automated pattern logging, image annotation, Excel-based result storage, and video generation of detected patterns over time.
Training Data
Fine-tuned on a custom dataset of 9,000 training images and 800 validation images, captured from diverse trading setups.
Key Strengths
- Real-time inference on screen-captured input, suitable for integration with live trading workflows.
- Optimized YOLOv8s architecture balances speed and accuracy for continuous monitoring.
Limitations
Detection accuracy can vary with chart style, screen resolution, and market conditions. Rapid fluctuations or patterns underrepresented in the training data may reduce reliability. Users should validate performance with historical and live data before deploying for trading decisions.
Example Detection
Developed by FODUU AI.
best for
- ·Real-time monitoring of stock chart patterns during trading hours
- ·Automated alerting for specific patterns like head-and-shoulders or triangles
- ·Integration into algorithmic trading systems for pattern-based decisions
FAQ
It detects six patterns: Head and shoulders bottom, Head and shoulders top, M_Head, StockLine, Triangle, and W_Bottom.
Use the OpenAI-compatible endpoint with your API key. Send an image and receive detected patterns as a response.
It expects an image, typically a screen capture of a stock market chart. The model processes the image and returns bounding boxes with pattern labels.
For customization inquiries, contact FODUU AI at [email protected].
We're benchmarking and onboarding Stock Market Pattern Detection YOLOv8s as a hosted, OpenAI-compatible API. Sign in for free credit and be ready when it lands, or tell us you want it and we'll prioritize it.