File size: 4,450 Bytes
98bfb69 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 | {
"version": "1.0.0",
"models_license": "CC-BY-4.0",
"base_url": "https://huggingface.co/JefersonPMS/agroforestry-inventory/resolve/main",
"read_this_first": [
"None of these models transfers to another site without loss. This is measured, not a",
"disclaimer: a coffee model scoring mAP 0.826 on its own validation set found 2 of 198",
"plants in held-out windows a few hundred metres away - recall 0.01 - while its precision",
"stayed at 1.00. It did not hallucinate; it went blind, and nothing in the output said so.",
"Before trusting a count on your imagery, check a sample by hand. The numbers below hold",
"for the site and the flight each model was trained on, and for nothing else."
],
"how_they_were_measured": {
"precision_and_recall": "Against hand-drawn labels, inside completed windows, matched one-to-one within 1 m. Only windows whose labels are entirely manual are used - comparing a model against labels it produced itself would be circular.",
"map50": "Training validation set, selected by k-means spatial stratification. Diagnostic only: a high mAP on a poorly distributed validation set is misleading, and misled us once.",
"position_error_m": "Median distance between the detected centroid and the label centroid, over matched pairs."
},
"models": [
{
"species": "coffee",
"title": "Arabica coffee (Coffea arabica)",
"file": "cafe.pth",
"classes": ["coffee"],
"site": "Raizes farm, Sao Paulo, Brazil",
"gsd_cm": 1.66,
"median_crown_m": 0.36,
"training_labels": 324,
"training_windows": 18,
"map50": 0.423,
"precision": 0.72,
"recall": 0.91,
"position_error_m": 0.03,
"detections_at_site": 9818,
"recommended_dedup_m": 0.3,
"where_it_fails": "The smallest crown in the set (0.36 m). Under closed canopy or heavy shading, neighbouring plants stop being separable. Outside the Raizes farm there is no guarantee at all.",
"sha256": "1e5ad88e96317bac64d454b566b20fad019a5819598d8a17e0e61f8d86395275",
"bytes": 120824288
},
{
"species": "avocado",
"title": "Avocado (Persea americana)",
"file": "abacate.pth",
"classes": ["avocado"],
"site": "Raizes farm, Sao Paulo, Brazil",
"gsd_cm": 1.66,
"median_crown_m": 1.52,
"training_labels": 120,
"training_windows": 27,
"map50": 0.773,
"precision": 0.69,
"recall": 0.94,
"position_error_m": 0.12,
"detections_at_site": 141,
"recommended_dedup_m": 0.7,
"where_it_fails": "Confuses other broad dark-leaved crowns when they share the scene. Trained on 120 labels only.",
"sha256": "5acb69c5c68da5df472d44277b748e3ae7f4bbaba2c529c6089f6f39d9577647",
"bytes": 120824288
},
{
"species": "pitaya",
"title": "Pitaya (Hylocereus spp.)",
"file": "pitaia.pth",
"classes": ["pitaya"],
"site": "Raizes farm, Sao Paulo, Brazil",
"gsd_cm": 1.66,
"median_crown_m": 0.97,
"training_labels": 100,
"training_windows": 22,
"map50": 0.863,
"precision": 0.89,
"recall": 0.98,
"position_error_m": 0.07,
"detections_at_site": 193,
"recommended_dedup_m": 0.4,
"where_it_fails": "The best model in the set, and the one with the fewest labels - what it has is the widest spatial distribution. The crop is trellised at regular spacing; performance on irregular planting was not measured.",
"sha256": "78c95933e2c4618c15d90d214fdcbc75ad015654ea759ff1c727719678bc19e8",
"bytes": 120824288
},
{
"species": "banana",
"title": "Banana (Musa spp.)",
"file": "banana.pth",
"classes": ["banana"],
"site": "Raizes farm, Sao Paulo, Brazil",
"gsd_cm": 1.66,
"median_crown_m": 3.11,
"training_labels": 133,
"training_windows": 14,
"map50": 0.817,
"precision": 0.59,
"recall": 0.93,
"position_error_m": 0.18,
"detections_at_site": 1567,
"recommended_dedup_m": 1.4,
"where_it_fails": "The worst precision in the set: 0.59. Banana grows in clumps, and the model often counts suckers of the same mat as separate individuals. If you need mats rather than leaves, this error is systematic and you will overcount.",
"sha256": "bc54d68beb4d2fcb080b3ba050ffe73fe3759c9da31cde9b5b5682ebadbfd922",
"bytes": 120824288
}
]
}
|