---
name: whiteboard-detector
version: 0.1.0
kind: onnx
tags:
  - yolo
  - detection
  - whiteboard
  - shapes
capabilities:
  - image_to_boxes
input:
  type: image
  formats:
    - png
    - jpeg
output:
  type: boxes
base_model: yolov8n
author: blockxaero
created: 1734048000
model_file: best.onnx
---

# Whiteboard Detector

YOLOv8-nano for whiteboard shape detection.

## Specs

- **Arch**: YOLOv8-nano
- **Format**: ONNX
- **Size**: ~12 MB
- **Input**: 640×640 RGB
- **Classes**: 30

## Classes

```
rectangle, rounded_rectangle, oval, circle, diamond, hexagon,
parallelogram, triangle, star, cloud, cylinder, stick_figure,
arrow_box, document_shape, database_icon, square, ellipse,
pentagon, cross, heart, lightning, banner, callout, bracket,
solid_arrow, dashed_arrow, bidirectional_arrow, dotted_line,
curved_arrow, curved_line
```

## Output

Tensor `[1, 34, 8400]` → post-process with NMS.

## License

BSL-1.1 © Block Xaero Inc.