| LibreSSD300 weights | |
| ----------------------- | |
| This product contains weights released by torchvision | |
| (https://github.com/pytorch/vision) at commit 336d36e8db990a905498c73933e35231876e28bc. | |
| Official checkpoint: ssd300_vgg16_coco-b556d3b4.pth | |
| Copyright (c) Soumith Chintala 2016 and torchvision contributors. | |
| The checkpoint has no separate per-object license file. Redistribution uses | |
| the releasing project's BSD-3-Clause license on an explicitly disclosed | |
| implied basis; this is not a publisher-confirmed checkpoint-specific grant. | |
| Torchvision warns that pretrained-model terms may derive from training data | |
| and users must determine permission for their use case. | |
| The SSD backbone was initialized from VGG-16 feature weights released by the | |
| Visual Geometry Group, University of Oxford, under Creative Commons | |
| Attribution 4.0 International: | |
| https://www.robots.ox.ac.uk/~vgg/research/very_deep/ | |
| https://creativecommons.org/licenses/by/4.0/ | |
| Creators: Karen Simonyan and Andrew Zisserman. Work: "Very Deep Convolutional | |
| Networks for Large-Scale Image Recognition," ICLR 2015. | |
| Torchvision modified the VGG graph for SSD and trained the detector on COCO. | |
| LibreYOLO preserves the learned tensors unchanged and adds checkpoint metadata | |
| only. The Oxford attribution records initialization lineage and does not claim | |
| that Oxford licensed the complete SSD checkpoint. | |