LibreSSD300 -- third-party attribution
======================================

Source: https://github.com/pytorch/vision
Commit: 336d36e8db990a905498c73933e35231876e28bc (torchvision v0.26.0)
License: BSD-3-Clause
Copyright (c) Soumith Chintala 2016

The native inference graph in nn.py is derived from torchvision's SSD300,
VGG feature extractor, default-box generator, box coder, detection transform,
and postprocessing utilities. LibreYOLO retains the upstream state-dict layout
and adds its own factory, checkpoint metadata, validation, contiguous COCO-80
class mapping, and export/backend integration. MultiBox matching,
hard-negative mining, and training losses are not included.

The official COCO checkpoint is not distributed in the LibreYOLO source tree.
Its separate LibreYOLO Hugging Face mirror uses BSD-3-Clause on an explicitly
disclosed implied basis: the checkpoint was released by the BSD-licensed
project but does not carry a per-object license file. Torchvision's README
warns that pretrained-model terms may derive from training data and that users
remain responsible for determining permission for their use case. The mirror
ships this BSD text and repeats that caveat; it must not be described as an
explicit checkpoint-specific grant.

VGG-16 feature-weight lineage
-----------------------------

The torchvision SSD300 training recipe initializes the detector backbone from
VGG-16 feature weights released by the Visual Geometry Group, University of
Oxford. The model was created by Karen Simonyan and Andrew Zisserman and is
described in "Very Deep Convolutional Networks for Large-Scale Image
Recognition" (ICLR 2015).

Source: https://www.robots.ox.ac.uk/~vgg/research/very_deep/
License: Creative Commons Attribution 4.0 International
License URL: https://creativecommons.org/licenses/by/4.0/

Changes and lineage: torchvision used the Oxford features-only weights to
initialize the VGG backbone, modified the architecture for SSD, and trained
the resulting detector on COCO. LibreYOLO preserves the released SSD learned
tensors unchanged and adds checkpoint metadata only. This attribution records
the VGG initialization lineage and does not claim that Oxford licensed the
complete SSD checkpoint.

BSD 3-Clause License

Copyright (c) Soumith Chintala 2016,

All rights reserved.

Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are met:

* Redistributions of source code must retain the above copyright notice, this
  list of conditions and the following disclaimer.

* Redistributions in binary form must reproduce the above copyright notice,
  this list of conditions and the following disclaimer in the documentation
  and/or other materials provided with the distribution.

* Neither the name of the copyright holder nor the names of its contributors
  may be used to endorse or promote products derived from this software without
  specific prior written permission.

THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE
FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
