How it works
Complete the job in three clear steps
- Choose a JPG, PNG, WebP, or BMP image from this device.
- Run the lightweight U2-Net foreground model locally in your browser.
- Review the edges and download the generated PNG.
Private on-device image AI
Keep the main subject sharp while blurring the detected photo background locally at full resolution.
Loading tool workspace…
How it works
Important
Keep an untouched copy of important source files. Browser processing depends on available device memory, and you should inspect every downloaded result before deleting the original.
Practical details
Keep the main subject sharp while blurring the detected photo background locally at full resolution.
The model can misclassify hair, glass, shadows, gaps, and similarly colored regions. Inspect the subject boundary and every person before publishing the result.
The lightweight general-object model is self-hosted and cached by the browser. Its first run downloads roughly 4.6 MB; later runs can reuse the local cache.
Questions
No. The browser downloads Lucaden's self-hosted U2-Net model, then runs inference against the image locally with ONNX Runtime Web.
The model can misclassify hair, glass, shadows, gaps, and similarly colored regions. Inspect the subject boundary and every person before publishing the result.
Open-source foundations
Lucaden documents its standards sources, architectural references, and permissive runtime dependencies while keeping this workspace consistent with the Lucaden interface.