tsumugi404/yolo26_flowchart_component

YOLO26 Flowchart Component Detection Model

This repository contains a YOLO26 object detection model fine-tuned for flowchart and block-diagram component detection.

The model is designed to detect common flowchart/diagram elements such as text regions, arrows, arrow start points, arrow end points, terminators, data blocks, process blocks, decision blocks, and connections.

Model Details

• Model architecture: YOLO26
• Task: Object Detection
• Domain: Flowcharts, block diagrams, document diagrams
• Framework: Ultralytics YOLO
• Input: Diagram / flowchart images
• Output: Bounding boxes and class labels for diagram components

Classes

The model predicts the following classes:

0: text
1: arrow
2: terminator
3: data
4: process
5: decision
6: connection
7: arrow_start
8: arrow_end

Additional annotations were manually created by the model author. These annotations were used to adapt the model to flowchart reconstruction and arrow endpoint detection.https://huggingface.co/datasets/tsumugi404/flowchat_components