# From vision-language model to edge detector

> Use a large vision-language model to label images, then train a compact detector that runs in real time on edge hardware.

Source: https://aibyos.com/research/edge-vision

Research · Vision

# From vision-language model to edge detector

Use a large vision-language model to label images, then train a compact detector that runs in real time on edge hardware.

**<20 ms**per frame on edge devices

**no cloud**inference on site

**VLM**auto-labelling cuts manual work

## Pipeline

1.  Auto-label images with a vision-language model
2.  Human review of a sampled subset
3.  Train a YOLO-class detector
4.  Quantise and compile for the target device
5.  Feedback loop from operators

Figures show the typical order of magnitude for this approach compared with calling a large general-purpose model. Actual results depend on the task and data; we measure them on your data during the baseline phase.

[Discuss a project like this](https://aibyos.com/contact)

## Typical tooling

-   Qwen-VL / Florence
-   YOLO
-   TensorRT / ONNX
-   Jetson

## Related

[All research](https://aibyos.com/research)[Sovereign GPU compute](https://aibyos.com/gpu)[Fine-tuning](https://aibyos.com/services/fine-tuning)