YOLOv8n Table Extraction Model
| Property | Value |
|---|---|
| Author | keremberke |
| Framework | PyTorch / Ultralytics |
| Accuracy | 96.71% mAP@0.5 |
| Downloads | 4,808 |
What is yolov8n-table-extraction?
The yolov8n-table-extraction is a specialized implementation of the YOLOv8 nano architecture designed specifically for detecting and localizing tables in documents. It can identify both bordered and borderless tables with high precision, making it particularly useful for document processing and data extraction workflows.
Implementation Details
Built using the Ultralytics YOLOv8 framework (version 8.0.21), this model leverages state-of-the-art object detection capabilities to identify table structures. It utilizes confidence thresholds and IoU parameters for optimal detection performance.
- Supports detection of both bordered and borderless table types
- Implements NMS with configurable confidence and IoU thresholds
- Achieves 96.71% mAP@0.5 on validation datasets
- Maximum detection capacity of 1000 instances per image
Core Capabilities
- Accurate table boundary detection
- Support for multiple table instances in single document
- Easy integration with Python applications
- Optimized for both accuracy and speed
Frequently Asked Questions
Q: What makes this model unique?
This model specializes in table detection with exceptional accuracy (96.71% mAP@0.5), making it particularly effective for document processing tasks. Its ability to detect both bordered and borderless tables sets it apart from general-purpose object detection models.
Q: What are the recommended use cases?
The model is ideal for document processing systems, automated data extraction from tables, digital document management systems, and any application requiring automated table detection in scanned documents or digital files.
