This repository's own source code is licensed under Apache-2.0. It does not contain any third-party model checkpoints. Install-time downloads remain subject to the terms published by their respective providers.
| Component | Use | License / status |
|---|---|---|
| LayerD | fallback layer decomposition | Apache-2.0 |
| PyTorch / torchvision | model inference and Mask R-CNN | BSD-style licenses |
| FastAPI, Uvicorn, Pillow, NumPy, SciPy, scikit-image, scikit-learn, Kornia, einops, timm, Transformers | application and ML runtime | Their upstream licenses; installed separately by the package manager |
| Tesseract | optional local OCR executable | Apache-2.0 |
uv.lock records the resolved Python dependency set. A distributor must retain
the notices and license texts required by each resolved dependency.
| Checkpoint | Upstream license / status | Release policy |
|---|---|---|
facebook/sam-vit-base |
Apache-2.0 | Downloaded from upstream; not committed |
| torchvision Mask R-CNN weights | BSD-style PyTorch ecosystem terms | Downloaded from upstream; not committed |
| Big-LaMa | Apache-2.0 upstream project | Downloaded from upstream; not committed |
gradio/Modnet ONNX checkpoint |
License metadata was not confirmed during this release audit | Explicit opt-in only; do not redistribute without verifying terms |
ZhengPeng7/BiRefNet_HR-matting checkpoint |
License metadata was not confirmed during this release audit | Explicit opt-in only; do not redistribute without verifying terms |
The four files in assets/generated_v2_sources/ were generated for this project
and are documented in that directory's README. They are included as benchmark
inputs; no third-party brand assets, watermarks, or intentionally readable text
were used.