edgetam_deploy
Copyright (c) Meta Platforms, Inc. and affiliates.

This product includes software developed by Meta Platforms, Inc., extracted from the
EdgeTAM project (https://github.com/facebookresearch/EdgeTAM) and reduced to its
single-image inference path. It is licensed under the Apache License, Version 2.0;
see the LICENSE file in this directory, which is byte-identical to the one published
there.

Like SAM 2 and unlike SAM 3, EdgeTAM's published checkpoint is Apache-2.0 as well --
Meta licenses the code and the model weights under the same terms. Nothing in this
family carries a separate weights licence.

EdgeTAM's image trunk is RepViT-M1, which upstream does not carry: it builds it through
the timm library at construction time. The modules are carried here instead, extracted
from that library, whose copyright notice travels with them:

  - PyTorch Image Models (timm)
    Copyright 2019 Ross Wightman
    https://github.com/huggingface/pytorch-image-models
    Apache License 2.0

  timm's own implementation is in turn adapted from the RepViT authors' release:

  - RepViT: Revisiting Mobile CNN From ViT Perspective
    https://github.com/jameslahm/RepViT
    Apache License 2.0

NOT INCLUDED
------------
The video tracker is not part of this package: memory attention, the memory encoder,
the 2-D Spatial Perceiver that is EdgeTAM's own contribution, object pointers and the
frame bookkeeping that drives them are all left behind, along with the training code
and the automatic mask generator. See PROVENANCE.md for the full list and for why.
