IBBI-bio/bark-ambrosia-beetle-benchmark

Bark and Ambrosia Beetle Detection Benchmark

v2.0.0 · 14,491 images · 175 species · 21 tribes · 70 genera

A specimen-disjoint, species-level object-detection benchmark for bark and ambrosia beetles (Coleoptera: Curculionidae: Scolytinae and Platypodinae), derived from the Bark and Ambrosia Gallery. Species determinations are made or reviewed by taxonomists; individual specimens carry bounding boxes.

Why this benchmark exists

Bark and ambrosia beetles are among the most damaging invasive forest insects worldwide, and the regulatory decisions they trigger — quarantine, eradication, port interception — are made at the level of species. Most wood-boring insects intercepted at ports are never identified below family, because the specialists who can separate morphologically near-identical congeners are few.

Automated identification is the obvious remedy, but the four things a model needs have not previously co-occurred in one dataset: expert determination, per-specimen localization, coverage of how these beetles are actually photographed, and a disjoint evaluation standard. This benchmark supplies all four.

What it measures that other benchmarks do not

This is not a generic detection benchmark with insects in it. It is built to separate failure modes that ordinary mAP conflates.

Detection and identification are scored separately. A detector here does two jobs — find the beetle, then name it. The benchmark reports detection recall, species accuracy given detection, and their product, so you can see which capability is failing. On this data they fail for different reasons and on different species: detection is close to solved (≈93% recall, median IoU 0.98) while identification is not (≈60% of found specimens named correctly). A single mAP number hides that entirely.

Fine-grained confusion is structured, not random. Species in the same genus are visually near-identical. Misidentifications land in the correct genus about 6× more often than chance, and consistently misidentified species get absorbed into abundant look-alike labels. The full confusion matrix and the taxonomy for every species are included so this structure can be analysed rather than averaged away.

Presentation is a controlled variable. Specimens appear both as isolated individuals and on dense plates carrying tens or hundreds of beetles at once — the bulk output of ethanol-baited survey traps. Per-species imaging density is recorded, so you can ask whether a model fails on a taxon or on a photograph.

Taxonomic novelty is stratified. 110 species are held out entirely and banded by distance from the training set (within-genus, within-tribe, outside-tribe), which separates "can it find an unfamiliar beetle" from "can it name one".

Domain shift is measurable. A separate iNaturalist split contains field photographs of species that are in the training set, so photographic domain transfer is isolated from taxonomic transfer.

Annotation budget is a controlled axis. Three data-allocation regimes × seventeen budgets let you ask how a fixed number of expert-annotated images is best spent — breadth across species versus depth within them.

Splits

Split	Images	Scored annotations	Crowd regions	Purpose
train	6,096	53,040	0	65 trainable species
iid_test	620	650	4,367	held-out specimens, trained species
inat_test	74	80	8	field photographs, domain shift
semantic_ood	7,701	46,924	0	110 unseen species, banded by distance

No image or specimen appears in more than one split.

iid_test and inat_test are balanced to 10 scored annotations per species, so no species dominates the mean and per-species AP is comparable across the label set.

Crowd regions — read before evaluating

Because the test splits are balanced to 10 annotations per species, and some of those annotations were drawn from dense plates, other specimens on the same image are real but not scored. They are marked iscrowd=1 with their true category_id.

pycocotools handles this correctly with no changes: crowd regions are excluded from the recall denominator, and a detection landing on one is neither rewarded nor penalised.

Custom evaluation code must honour iscrowd. Treating crowd regions as ordinary targets inflates the iid_test ground truth from 650 to 5,017 and collapses recall. A crowd-aware reference evaluator is included at evaluation/evaluate.py, alongside evaluation/aggregate.py. Verify any evaluator before use: grep -c "_n_real_gt" evaluate.py should return 4 or more. The v1.0.0 evaluator is not crowd-aware and must not be used with v2 annotations.

train and semantic_ood are exhaustively annotated and contain no crowd regions.

Reference results

Five architectures at full training data (YOLOv8x, YOLOv10x, YOLO11x, YOLO12x, RT-DETR-X; mean of three seeds):

Metric	Range	What it tells you
iid_test AP@0.5	0.504 – 0.558	joint detect-and-name
Detection recall	0.926 – 0.936	finding the beetle: close to solved
Species accuracy given detection	0.581 – 0.626	naming it: the bottleneck
End-to-end species recall	0.538 – 0.586	what a deployed tool delivers
inat_test AP@0.5	15–35% of IID	domain transfer is poor
Novelty AUROC (confidence)	0.54 – 0.71	no usable abstention signal

The gap between detection recall and species accuracy is the headline: models lose more than five times as much performance to misidentification as to localization. Architecture choice barely moves this — per-species AP vectors correlate at ρ ≈ 0.96 across architectures, so the ceiling is a property of the data and the task, not the backbone.

Three capabilities are unsolved and are what this benchmark is for measuring: discriminating congeners, detecting specimens in dense multi-specimen frames, and recognising a species outside the label set.

Format

COCO detection JSON. Each image record carries source_image_uuid (links to the platform), width, height, and full licence metadata. Each annotation carries source_record_id, category_id, absolute-pixel bbox, area, iscrowd, and depicts_valid_name_id.

from pycocotools.coco import COCO
coco = COCO("detection/annotations_coco/iid_test.json")
real = [a for a in coco.loadAnns(coco.getAnnIds()) if not a["iscrowd"]]
print(len(real))   # 650

Licensing

Images carry mixed Creative Commons licences and are redistributed unmodified. There is no single dataset-level licence.

Licence	Images	Version stated
CC BY-NC 4.0	11,432	yes
CC BY	2,358	no
CC0 1.0	293	yes
CC BY-SA	192	no
CC BY-NC-SA	113	no
CC BY-NC	103	no
Total	14,491

Per-image licence, photographer and holding institution are in image_licences.csv. Each COCO image record carries license_name, license_url, license_verbatim and license_versioned.

2,766 images (19.1%) come from sources that state no licence version. These are recorded as supplied with license_versioned: false and are not upgraded to 4.0. Users should satisfy the terms of the most restrictive plausible version. This concentrates in inat_test, where 61 of 74 images are unversioned CC BY-NC.

No no-derivatives (ND) images are present in any split. Images are redistributed unmodified as a Collection; ShareAlike obligations attach to adaptations rather than collections, so the 305 SA-family images do not impose a licence on the dataset as a whole.

Annotations, splits and metadata are released under CC BY 4.0.

Citation

@article{marais_bark_ambrosia_gallery,
  title   = {The Bark and Ambrosia Gallery: an expert-curated image resource,
             curation platform, and machine-learning benchmark for
             Scolytinae and Platypodinae},
  author  = {Marais, Christopher and Schuster, Layla A. and others},
  journal = {Database},
  note    = {in review}
}

See CHANGELOG.md for what changed in v2.0.0.