Views
No views yet
LibreSAM promptable-segmentation family: a
1,371,418-parameter CNN that segments the object inside a box-defined region of
interest. It is small enough to run in-sensor on the Sony IMX500 and is the
smallest promptable segmenter LibreYOLO ships, sitting below MobileSAM.1from libreyolo import LibreSAM
2
3model = LibreSAM("picosam3")
4result = model.predict("image.jpg", bboxes=[100, 100, 400, 500])
5result.show()bboxes= is the only supported prompt. Point, text, mask, multimask and
segment-everything modes are not part of the upstream model contract and raise
a clear error; use LibreSAM("sam2") or LibreSAM("sam3") for those.| Protocol | mIoU |
|---|---|
| Crop space, 96x96 (upstream evaluation protocol) | 0.692 |
Full image, end-to-end via LibrePicoSAM3.predict() | 0.697 |
| Model | Params | mIoU | CPU latency |
|---|---|---|---|
| LibrePicoSAM3 | 1.37M | 0.691 | 8.6 ms/prompt |
| LibreMobileSAM | 10.13M | 0.800 | 407 ms/prompt |
1b03949e43472953bb0021685c7fc3f5fdf48fde. The LibreYOLO
implementation is native and produces bit-identical outputs to upstream
(max abs diff 0.0 on a seeded FP32 batch).af49e4322b6b7cf448499fee5c073d4576f59444, file
PicoSAM3_SAM3_student_best.pt. Re-wrapped into the LibreYOLO checkpoint
schema (v1) with provenance metadata; tensor values are unchanged.PicoSAM3_student_epoch1.pt and PicoSAM3_epoch1.pt files in the
upstream weights repo contain the older PicoSAM2 architecture (output_head.*)
and do not load as PicoSAM3. LibreYOLO ships the best artifact and rejects the
epoch-1 files explicitly.1@article{picosam3_2026,
2 title={PicoSAM3: Real-Time In-Sensor Region-of-Interest Segmentation},
3 journal={IEEE Sensors Journal},
4 year={2026}
5}