LibreFCOSr50

FCOS with a ResNet-50 FPN backbone, repackaged for LibreYOLO.

from libreyolo import LibreYOLO

model = LibreYOLO("LibreFCOSr50.pt")
results = model.predict("image.jpg")

Source

Derived from pytorch/vision at commit 336d36e8db990a905498c73933e35231876e28bc. Copyright (c) Soumith Chintala 2016 and torchvision contributors. The source implementation is BSD-3-Clause.

Official checkpoint: fcos_resnet50_fpn_coco-99b0c9b7.pth SHA-256: 99b0c9b7cfb1527d782db86b91d207f00547c792fb4103fc612b651d0a07b9e7 Published COCO val2017 box mAP: 39.2.

Modifications

Checkpoint metadata was added for LibreYOLO's v1.0 schema. Learned tensors and state-dict keys are unchanged. The native LibreYOLO graph loads all 319 official state entries strictly and matches the pinned source at raw heads, anchors, preprocessing, and final detections. See weights/convert_fcos_weights.py in the LibreYOLO source repository.

Benchmarks

Independent accuracy and speed benchmarks: visionanalysis.org/model/fcos-r50

License

The checkpoint publisher did not attach a separate per-object license file. This mirror applies the releasing project's BSD-3-Clause license on an implied, not publisher-confirmed, basis. Torchvision warns that pretrained models may have their own licenses or terms derived from training data and that users must determine whether they have permission for their use case. COCO annotations are CC BY 4.0; source images retain their individual Flickr terms. See LICENSE and NOTICE.