IBBI YOLO11x arthropod detector
A single-class detector (arthropod) that finds any arthropod in any imaging setting: lab dishes and scanners,
light traps, sticky cards and pitfall trays, camera traps over vegetation, and field photographs. It is stage 1 of the
ibbi identification pipeline; the hierarchical classifier names what it finds.
import ibbi
det = ibbi.create_model("arthropod_detector")
boxes = det.predict("trap.jpg", conf=det.operating_conf) # high-precision operating point (0.70)
pipe = ibbi.create_pipeline() # detector + hierarchical classifier
Model
| Architecture | YOLO11x (Ultralytics), COCO-pretrained start |
| Input size | 1024 px |
| Parameters | 56.9 M |
| Training data | IBBI arthropod detection corpus v3: 307,421 images and 889,283 boxes from 14 datasets in 5 environment categories; train split 239,477 images, square-root source balancing |
| Training | 20 epochs of the balanced mix, batch 32, AdamW lr 1e-3, fp32, 4×B200 |
| Checkpoint | epoch 20, chosen before testing by the highest recall at ≤ 0.2 false alarms per image on the corpus validation sample |
| Operating confidence | 0.70: at most 0.2 false alarms per image on exhaustively labelled validation images |
Benchmark results (class-agnostic, through ibbi.Evaluator)
Crowd-aware scoring on the beetle benchmark, every box treated as "object". Caveat: about 99% of the benchmark images are also in the detector corpus (the Bark and Ambrosia Gallery is one of its sources), many of them in its training split, so these recall numbers are optimistic for the benchmark images.
| Split | AP@[.50:.95] | AP@0.5 | AR@100 | max recall (IoU 0.5) | recall at op. conf | precision at op. conf | false alarms / image |
|---|---|---|---|---|---|---|---|
| iid_test | 0.744 | 0.774 | 0.909 | 0.932 | 0.922 | 0.435 | 1.25 |
| inat_test | 0.932 | 0.996 | 0.945 | 1.000 | 0.963 | 0.987 | 0.01 |
| semantic_ood | 0.632 | 0.670 | 0.763 | 0.797 | 0.784 | 0.795 | 1.23 |
Operating confidence: 0.7. It was set on the detector corpus validation sample; on the benchmark it gives the false-alarm rates above, so choose conf for your own images when false alarms matter.
Licence
Weights: AGPL-3.0 (Ultralytics); see LICENSE.
The training data include sources whose terms forbid commercial use: iNaturalist 2017 (non-commercial research and education only) and IP102 (academic use only); Pest24's original terms could not be verified. The remaining sources are CC BY 4.0, CC BY 2.0 (Open Images), MIT (AMI traps) or the project's own images. Whether model weights inherit the terms of their training data is legally unsettled; these weights are released for non-commercial research use, and commercial use is not supported without permission from those data owners.
Citation
@dataset{marais_bark_ambrosia_benchmark_2026,
title = {Bark and Ambrosia Beetle Detection Benchmark},
author = {Marais, G. Christopher and Schuster, Layla A. and Johnson, Andrew J. and Kuo, Eric and Dias, Raquel and Hulcr, Jiri},
year = {2026},
publisher = {Zenodo},
version = {2.0.1},
doi = {10.5281/zenodo.22695714}
}
The ibbi package: https://github.com/ChristopherMarais/IBBI