nrl-ai / anylearning-labeling-models

AnyLearning labeling models

Versioned ONNX model bundles used by AnyLearning for local, prompt-guided image segmentation.

This repository contains the five models currently offered by AnyLearning:

Each ZIP contains a small AnyLearning model configuration and one encoder plus one decoder ONNX model. MANIFEST.json records the exact source revision, archive size, checksum, and expected members.

Provenance

The files are byte-for-byte mirrors of ONNX exports published by Viet-Anh Nguyen:

• SAM 2 bundles: source revision 071f58077599431edd0e5d2ac52ecca4c78f1cab from vietanhdev/segment-anything-2-onnx-models .
• MobileSAM bundle: source revision 9effc01a9e135621d710d49159f1ffb0b6f724dc from vietanhdev/segment-anything-onnx-models .

Original model projects:

• Segment Anything 2
• Segment Anything
• MobileSAM

These mirrors do not change the weights or model graphs.

Secure and reproducible download

Pin a repository revision and verify the SHA-256 value from MANIFEST.json before extracting or loading a model. Consumers should reject absolute paths, parent traversal, links, unexpected archive members, and files exceeding their configured size limits.

from hashlib import sha256
from pathlib import Path

from huggingface_hub import hf_hub_download

path = Path(
    hf_hub_download(
        repo_id="nrl-ai/anylearning-labeling-models",
        filename="sam2_hiera_tiny.zip",
        revision="v1.0.0",
    )
)

expected = "7454c3afd835b2acaad863afe3acb11f4e4af039c96e886989ad3873a338e1ec"
assert sha256(path.read_bytes()).hexdigest() == expected

AnyLearning performs the full image encoding, prompt conversion, mask decoding, and editable-shape conversion. These archives are not standalone applications.

Intended use and limitations

• Intended for interactive point/rectangle-prompt segmentation in AnyLearning.
• Results require human review before becoming dataset labels.
• Quality and latency vary with image content, hardware, execution provider, and model size.
• These models can reproduce biases and limitations of their original training data.
• Do not use segmentation output as the sole basis for safety-critical, medical, legal, or similarly consequential decisions.

License

The model code and weights are distributed under Apache License 2.0 by their respective upstream projects. See LICENSES.md for source and attribution links.

Citation

For SAM 2, cite:

@article{ravi2024sam2,
  title={SAM 2: Segment Anything in Images and Videos},
  author={Ravi, Nikhila and others},
  journal={arXiv:2408.00714},
  year={2024}
}

For SAM, cite:

@article{kirillov2023segment,
  title={Segment Anything},
  author={Kirillov, Alexander and others},
  journal={arXiv:2304.02643},
  year={2023}
}