guillherms/vision-architecture-analyzer-yolo11-detect

Vision Architecture Analyzer YOLO11 Detect - v3.0.0

Overview

This model detects cloud architecture components in AWS/Azure-style diagrams.

It outputs bounding boxes with:

• class id
• class name
• confidence score

Training Details

• Base model: YOLO11s
• Image size: 1280
• Dataset: STRIDE Architecture Components (AWS & Azure)
• Augmentations: disabled geometric transforms to preserve diagram structure
• Run dir: train2

Artifacts

Version folder: v3.0.0/

Latest folder: latest/

Includes:

• best.pt (also uploaded at repo root)
• last.pt (also uploaded at repo root)
• args.yaml (if available)
• results.csv (if available)
• training plots (if available)
• dataset data.yaml
• classes.txt

Usage Example

from ultralytics import YOLO

model = YOLO("best.pt")
results = model.predict("image.png", imgsz=1280, conf=0.2)