Model zoo¶
spaCR ships a catalogue of trained models and fetches them on demand. Name a
key in a settings file — pathogen_model: toxoplasma_pv_v1 — and the model
is downloaded and checksum-verified the first time it is needed, or open
Model Zoo from the home screen to browse and install them.
Every published entry carries a SHA-256. An entry without one is refused rather than installed, because a truncated or substituted checkpoint cannot be told from the real one.
Model |
Training data |
Hold-out performance |
|---|---|---|
|
anti-Toxoplasma-biotin and DsRed PV lumen; 115 images, 1 dataset |
F1 0.867 against 0.713 for stock cpsam, at IoU 0.5 |
|
crystal violet plaque wells; 184 wells from 3 datasets, 95 in-house and 89 literature |
F1 0.856 in-domain; 0.806 on literature (3-fold cross-validated, SD 0.020) |
|
whole-plate and multi-well crystal violet images; 562 images from 1 dataset, 190 of them with no well in them |
mAP50 0.993, mAP50-95 0.886, precision and recall both 0.987 |
Models are hosted on their author’s own Hugging Face account, so contributing
one does not mean handing write access to anyone else’s.
spacr.model_zoo’s publish_model performs the upload and prints the
catalogue row to add.
Per-model detail¶
Toxoplasma PV v1¶
Architecture. Cellpose-SAM (cpsam_v2)
Trained on. Toxoplasma tachyzoite parasitophorous vacuoles stained with goat anti-Toxoplasma-biotin, and tachyzoites expressing DsRed in the PV lumen. 115 pairs (104 train / 11 test), 100 epochs, base cpsam_v2
Measured. F1 0.867 against 0.713 for stock cpsam, at IoU 0.5
F1 0.867 at IoU 0.5 against 0.713 for stock cpsam; AJI 0.808 against 0.426
accuracy falls sharply above IoU 0.8 – suited to counting and area rather than precise morphometry
Published as einarolafsson/toxoplasma-pv-segmentation-cpsam, as cpsam_v2_toxo_r2.
SHA-256 182d8cf6b32c7b9ef2917c85870d188486e5e119f05e9c5c1f07652f6859f2d0.
Toxoplasma Plaque v1¶
Architecture. Cellpose-SAM (cpsam)
Trained on. Toxoplasma gondii plaque assays; round 3, evaluated in-domain (NAS) and against a literature generalisation set
Measured. F1 0.856 in-domain; 0.806 on literature (3-fold cross-validated, SD 0.020)
F1 0.856 in-domain and 0.806 on the literature set (3-fold cross-validated, SD 0.020), against 0.718 for round 1
round 3 trades precision (0.939 down to 0.858) for recall (0.631 up to 0.811) on the literature set, which is the right direction for a counting assay
Published as einarolafsson/toxoplasma-plaque-segmentation-cpsam, as cpsam_plaque_r3.
SHA-256 eeecd2d6cd5cbb4dddee71564d5f460d26bb07ac125e0b494b7502fea4292d5d.
Toxoplasma Plaque Well Detector v1¶
Architecture. YOLO11n
Trained on. whole-plate and multi-well Toxoplasma plaque-assay images; yolo11n base, 150 epochs, batch 16, imgsz 640
Measured. mAP50 0.993, mAP50-95 0.886, precision and recall both 0.987
mAP50 0.993, mAP50-95 0.886, precision and recall both 0.987
locates WELLS, not plaques; it is the front half of a two-stage pipeline with toxoplasma_plaque_v1, and the well it finds also gives the diameter that makes areas comparable across microscopes
Published as einarolafsson/toxoplasma-plaque-well-detector-yolo11, as yolo_welldetect_v3.pt.
SHA-256 b826058754fb5d4df36c3a7283aac049015cbb044b5ef096c55d19f37172a50c.