Standing on Open Shoulders: Regional Fine-Tuning of Open-Weight Camera-Trap Models for UK Wildlife with Modest Data and Compute

Paul Fergus

cs.CVOct 23, 2026

Standing on Open Shoulders: Regional Fine-Tuning of Open-Weight Camera-Trap Models for UK Wildlife with Modest Data and Compute

Authors: Paul Fergus

Organizations: Liverpool John Moores University, Byrom Street, Liverpool L3 3AF, United Kingdom

Abstract

Camera-trap species classification is dominated by large models from well-resourced organisations, which often lack the species-level distinctions regional conservation needs, such as native red versus invasive grey squirrels in the UK. We ask whether open weights let a small team with a modest dataset and a single GPU take a real part in AI development. We fine-tuned the open DeepFaune classifier on 28 UK classes, paired it with the unmodified MegaDetector v6, and compared the resulting DeepFaune-UK v1.2 with DeepFaune v1.5.0 and SpeciesNet on 20,212 hand-verified MammalWeb citizen-science images, after removing sequences sharing any of 1,039 images with our training archive. DeepFaune-UK achieved a strict image-level accuracy of 88.8%, including empty frames, against 56.5% and 59.0% for the baselines, but at DeepFaune's coarser label resolution DeepFaune v1.5.0 was more accurate, though not significantly under contributor resampling (91.4% vs. 88.8%). As DeepFaune v1.5.0 uses a newer DINOv3-based classifier, we applied the identical recipe, without new data, to that classifier. The resulting DeepFaune-UK v1.3 reached 90.6% strict accuracy, narrowed the like-for-like gap to 0.8 points and reduced roe deer misclassified as fallow deer by 69%. On 1,850 images from the 26 contributors with no overlapping images, the three DeepFaune-based models did not differ significantly (90.5%, 90.6%, 89.9% like-for-like). Fine-tuning on 10% of the training crops (about 150 per class) gave out-of-distribution accuracy similar to using all of them, and a DINOv2 run on this subset took about 22 minutes with under 4 GB of GPU memory. Open weights thus make regional model development feasible for small teams, but the largest like-for-like gain came from a newer base model trained by others, so this democratisation depends on large actors continuing to release open weights, under licences that may change.

Figures & tables

Figure 1: Overview of the study. (a) Model development: the master dataset is split by capture group, MegaDetector v6 detections matched to the annotations (with the annotated box for any animal the detector missed) provide the training crops, and the DeepFaune classifier is fine-tuned in two phases from either its DINOv2 (v1.4) or its DINOv3 (v1.5.0) checkpoint. (b) Inference and evaluation: MegaDetector v6 localises animals, people and vehicles; only animal crops are classified, and the image label is that of the most confident animal detection. DeepFaune-UK and the two baselines are scored on the MammalWeb out-of-distribution set, which plays no part in training, classifier selection or threshold tuning (Sections 3.3 and 3.5).

Model	Strict accuracy	Strict macro-F1	Like-for-like accuracy	Like-for-like macro-P	Like-for-like macro-R	Like-for-like macro-F1	Like-for-like weighted F1
DeepFaune-UK v1.2	0.888	0.778	0.888	0.844	0.866	0.850	0.928
DeepFaune-UK v1.3	0.906	0.786	0.906	0.844	0.880	0.857	0.942
DeepFaune v1.5.0	0.565	0.398	0.914	0.886	0.829	0.855	0.947
SpeciesNet 5.0.5	0.590	0.536	0.734	0.911	0.605	0.708	0.809

Table 1: Image-level performance on the OOD evaluation set (20,212 images, 14 species plus empty). DeepFaune-UK v1.3 is described in Section 4.3. Best value per column in bold; sequence- and contributor-clustered 95% bootstrap confidence intervals for accuracy and macro-F1 are given in Tables B1 and B2.

Species	n	DeepFaune-UK v1.2	DeepFaune-UK v1.3	DeepFaune v1.5.0	SpeciesNet 5.0.5
Roe deer	5,361	0.926	0.940	0.968	0.674
Rabbit	3,584	0.939	0.958	0.950	0.934
Grey squirrel	2,742	0.962	0.960	0.923	0.960
Badger	1,448	0.959	0.973	0.971	0.907
Red fox	920	0.924	0.946	0.945	0.897
Cat	843	0.763	0.840	0.954	0.928

Table 2: Like-for-like per-species F1 on the OOD evaluation set. n = number of images containing the species; best value per row in bold (ties at three decimals are all bolded; no bold where all models score 0). At this level any squirrel label counts as correct for either squirrel species, so the red-squirrel values (three images, all from one sequence) do not show that red squirrels were identified: at the strict species level, no model labelled any of them as red squirrel (F1 = 0 for all four).

Figure 2: Like-for-like per-species F1 on the OOD evaluation set for the four models. Species are ordered by the number of images containing them (in brackets); cells are shaded by F1, with values below 0.5 sharing the lightest shade. Muntjac and red squirrel have only three images each, each set from a single sequence, so their values are unreliable; the red-squirrel values reflect the shared squirrel group, and no model identified a red squirrel at the strict species level.

Figure 3: Confusion matrices of (a) DeepFaune-UK v1.2 and (b) DeepFaune-UK v1.3 on the OOD evaluation set at the strict species level (20,168 single-species and empty images; the 44 images with more than one species are excluded). Numbers are image counts, and shading shows each count as a share of its true class on a square-root scale. Columns to the right of the vertical rule are predicted labels outside the 14 OOD species: fallow deer, sheep, wild boar, pine marten, the remaining UK classes (pooled), person or vehicle, and empty.

Training fraction	Training crops	ID test macro-F1	OOD strict macro-F1	OOD like-for-like accuracy	OOD like-for-like macro-F1	Training time (min)	Peak GPU memory (GB)	GPU energy (kWh)
10%	4,270	0.990 $\pm$ 0.002	0.767 $\pm$ 0.008	0.887 $\pm$ 0.005	0.838 $\pm$ 0.007	21.8 $\pm$ 0.9	3.8	0.064 $\pm$ 0.006
25%	10,681	0.992 $\pm$ $<$ 0.001	0.773 $\pm$ 0.005	0.888 $\pm$ 0.010	0.843 $\pm$ 0.006	20.6 $\pm$ 3.9	3.8	0.071 $\pm$ 0.011
50%	21,360	0.990 $\pm$ 0.003	0.767 $\pm$ 0.009	0.882 $\pm$ 0.018	0.838 $\pm$ 0.009	30.8 $\pm$ 5.5	3.8	0.122 $\pm$ 0.021
100%	42,719	0.993 $\pm$ 0.002	0.767 $\pm$ 0.009	0.879 $\pm$ 0.005	0.839 $\pm$ 0.009	44.3 $\pm$ 7.6	3.8	0.192 $\pm$ 0.034

Table 3: Learning curve and resource use (mean

\pm

s.d. over three seeds; single NVIDIA RTX A6000). OOD scores are image-level on the evaluation set; detector outputs are identical for all runs.

Backbone	Training fraction	ID test macro-F1	OOD strict macro-F1	OOD like-for-like accuracy	OOD like-for-like macro-F1	Training time (min)	Peak GPU memory (GB)	GPU energy (kWh)
DINOv2	10%	0.990 $\pm$ 0.002	0.767 $\pm$ 0.008	0.887 $\pm$ 0.005	0.838 $\pm$ 0.007	21.8 $\pm$ 0.9	3.8	0.064 $\pm$ 0.006
DINOv2	100%	0.993 $\pm$ 0.002	0.767 $\pm$ 0.009	0.879 $\pm$ 0.005	0.839 $\pm$ 0.009	44.3 $\pm$ 7.6	3.8	0.192 $\pm$ 0.034
DINOv3	10%	0.991 $\pm$ 0.002	0.779 $\pm$ 0.005	0.901 $\pm$ 0.013	0.846 $\pm$ 0.005	45.6 $\pm$ 14.7	4.4	0.114 $\pm$ 0.014
DINOv3	100%	0.994 $\pm$ $<$ 0.001	0.786 $\pm$ 0.003	0.903 $\pm$ 0.003	0.857 $\pm$ 0.003	91.2 $\pm$ 20.6	4.4	0.394 $\pm$ 0.088

Table 4: Backbone comparison: DINOv2 (DeepFaune v1.4, as in Table 3) and DINOv3 (DeepFaune v1.5.0) classifiers fine-tuned with the identical recipe (mean

\pm

s.d. over three seeds).

Figure 4: Learning curve and resource footprint of regional fine-tuning (mean

\pm

s.d. over three seeds; small points show individual seeds). (a–c) OOD performance against the number of training crops: (a) like-for-like accuracy, (b) like-for-like macro-F1 and (c) strict-species macro-F1; in (a) and (b) the dashed line marks DeepFaune v1.5.0, and SpeciesNet (like-for-like accuracy 0.734, macro-F1 0.708) lies below the plotted range; both baselines fall well below the range of (c) (strict macro-F1 0.398 and 0.536). (d–f) Training time, GPU energy and peak GPU memory per run on a single NVIDIA RTX A6000; the dotted line in (f) marks the 8 GB memory of a typical consumer GPU. Filled circles: DINOv2 backbone (DeepFaune-UK v1.2 recipe); open squares: DINOv3 backbone (DeepFaune-UK v1.3), trained at 10% and 100% only and offset slightly for legibility.

Model	Strict accuracy	Like-for-like accuracy	95% CI, sequences	95% CI, contributors	Like-for-like macro-F1
DeepFaune-UK v1.2	0.903	0.905	0.887–0.922	0.868–0.932	0.931
DeepFaune-UK v1.3	0.906	0.906	0.888–0.923	0.863–0.934	0.904
DeepFaune v1.5.0	0.515	0.899	0.880–0.917	0.843–0.927	0.877
SpeciesNet 5.0.5	0.616	0.791	0.764–0.818	0.714–0.858	0.687

Table 5: Site-disjoint subset: 1,850 images from the 26 contributors none of whose images were found as exact copies in the master dataset (622 sequences, 12 species plus empty; macro-F1 over the 12 species present). Confidence intervals from 2,000 cluster-bootstrap resamples.

Appendix figures & tables8 assets

Supplementary material from the paper’s appendix.

Appendix

Figure A1: Species-level resolution. DeepFaune-UK v1.2 and v1.3 name the species, while DeepFaune v1.5.0 names only a coarser group: (a) grey squirrel, (b) rabbit, (c) woodpigeon and (d) carrion crow. ✓ exact species; $\approx$ correct group only; ✗ neither (wrong or too coarse). Tiles are cropped around the annotated animal for display. Images: MammalWeb, CC BY-SA 4.0.

Figure A2: Images on which neither baseline returns the correct harmonised group (some baseline labels are correct but too coarse, such as “cervidae family” or “animal”) while both DeepFaune-UK models name the exact species: (a) roe deer, (b) hedgehog, (c) horse and (d) dog. Symbols as in Figure A1.

Figure A3: Further images on which neither baseline returns the correct harmonised group while both DeepFaune-UK models name the exact species: (a) carrion crow, (b) rabbit, (c) woodpigeon and (d) pheasant. Symbols as in Figure A1.

Figure A4: Errors of DeepFaune-UK v1.2 corrected by v1.3: (a) roe deer labelled fallow deer, (b) dogs labelled sheep, (c) grey squirrel and (d) red fox labelled cat. Symbols as in Figure A1.

Figure A5: Characteristic failures of DeepFaune-UK: (a) roe deer labelled muntjac by v1.3; (b) a red squirrel labelled grey squirrel by both versions; (c) a grey squirrel detected below the confidence threshold and reported as empty; and (d) an empty frame in which a false detection (on tree bark) was labelled fallow deer by both versions. Symbols as in Figure A1.

Model	Accuracy	95% CI, sequences	95% CI, contributors	Macro-F1	95% CI, sequences	95% CI, contributors
DeepFaune-UK v1.2	0.888	0.880–0.895	0.869–0.910	0.778	0.766–0.914	0.754–0.905
DeepFaune-UK v1.3	0.906	0.899–0.913	0.889–0.924	0.786	0.775–0.923	0.768–0.919
DeepFaune v1.5.0	0.565	0.550–0.580	0.399–0.715	0.398	0.395–0.466	0.390–0.466
SpeciesNet 5.0.5	0.590	0.577–0.603	0.509–0.702	0.536	0.525–0.630	0.513–0.640

Appendix

Table B1: Point estimates and 95% cluster-bootstrap confidence intervals (2,000 resamples) at the strict species level on the OOD evaluation set, resampling whole sequences or whole contributors (Section 3.6). Macro-F1 intervals are asymmetric because muntjac and red squirrel, each with three images from one sequence, are absent from 35–37% of resamples and, when present, score an F1 of zero at this level for all models.

Model	Accuracy	95% CI, sequences	95% CI, contributors	Macro-F1	95% CI, sequences	95% CI, contributors
DeepFaune-UK v1.2	0.888	0.881–0.896	0.870–0.912	0.850	0.829–0.921	0.806–0.926
DeepFaune-UK v1.3	0.906	0.899–0.913	0.890–0.925	0.857	0.837–0.929	0.817–0.934
DeepFaune v1.5.0	0.914	0.909–0.920	0.902–0.922	0.855	0.829–0.930	0.818–0.931
SpeciesNet 5.0.5	0.734	0.720–0.748	0.631–0.815	0.708	0.679–0.778	0.667–0.779
DeepFaune-UK v1.2 (calibrated)	0.889	0.881–0.896	0.869–0.911	0.837	0.825–0.915	0.812–0.913
DeepFaune-UK v1.3 (calibrated)	0.909	0.902–0.915	0.892–0.927	0.858	0.839–0.931	0.822–0.937

Appendix

Table B2: Point estimates and 95% cluster-bootstrap confidence intervals (2,000 resamples) at the like-for-like level on the OOD evaluation set, resampling whole sequences or whole contributors (Section 3.6). Calibrated rows use thresholds selected on the master validation split by the procedure of Section 4.2 (v1.2: detector 0.50, classifier 0.30; v1.3: 0.50 and 0.20, Section 4.3).

Model	Strict accuracy	Strict macro-F1	Like-for-like accuracy	Like-for-like macro-F1
DeepFaune-UK v1.2	0.894	0.782	0.895	0.854
DeepFaune-UK v1.3	0.911	0.791	0.912	0.863
DeepFaune v1.5.0	0.552	0.399	0.918	0.859
SpeciesNet 5.0.5	0.605	0.548	0.747	0.719

Appendix

Table B3: Point estimates on the original 21,640-image set, before the removal of the 440 sequences that overlap with the master dataset (Section 3.5), for comparison with Table 1.

Image Classification

Image classification assigns semantic labels to images, enabling recognition of objects, scenes, conditions, and fine-grained categories.

See all 386 papers

Democratising Camera Trap AI: An Open-Source Model for Detecting UK Mammals
Jun 9, 2026Paul Fergus, Philip Stephens, Russell A. Hill +8You Only Look Once Object Detection
1Liverpool John Moores University, Byrom Street, Liverpool, L3 3AF, UK. · 2Durham University, Stockton Road, Durham, DH1 3LE. · 3MammalWeb, c/o Durham Wildlife Trust, Houghton-le-Spring, DH4 6PU +6
DeepForestVisionV2: Ecology-Driven Taxonomy Expansion for Camera-Trap Monitoring in African Tropical Forests
Jun 18, 2026Hugo Magaldi, Theau d'Audiffret, Etienne Francois Akomo-Okoue +24Fine-Grained Image Classification Wildlife Monitoring
UMR7206 Eco-Anthropologie, MNHN, Paris, France; One Forest Vision initiative · Centre National de la Recherche Scientifique et Technologique, Libreville, Gabon · Institut de Recherche en Ecologie Tropicale, Libreville, Gabon +13
Camera trap classification with deep learning under ground truth uncertainty
Aug 31, 2026Leonard Hockerts, Peter S. Stewart, Sarthak Arora +1Learning with Noisy Labels Image Classification
School of Mathematics & Statistics, University of Glasgow, Glasgow, G12 8QQ,2026 United Kingdom. · 2Collaborative Earth, Berkeley, California, United States

Jun 9, 2026J/K move · Enter open · S save

Democratising Camera Trap AI: An Open-Source Model for Detecting UK Mammals

Paul Fergus, Philip Stephens, Russell A. Hill, Lee Oliver+7

1Liverpool John Moores University, Byrom Street, Liverpool, L3 3AF, UK. · 2Durham University, Stockton Road, Durham, DH1 3LE. · 3MammalWeb, c/o Durham Wildlife Trust, Houghton-le-Spring, DH4 6PU+6

You Only Look Once Object Detection Wildlife Monitoring

Open paper AskPDF arXiv

Camera traps have become a cornerstone of biodiversity monitoring, but the artificial intelligence that turns vast quantities of images into usable ecological data is often locked behind commercial platforms or trained on fauna that does not match that of the British Isles. In an attempt to remove barriers and increase uptake, we release an open-source object detection model for 31 classes, 28 common UK mammal and bird species, plus utility classes for humans, calibration poles, and vehicles, drawn from a curated dataset of 48,165 labelled instances assembled from multiple sites over a decade of operational deployment through Conservation AI and its successor, Trap Tracker. The model, a YOLO26x detector trained and tested on an 80/10/10 class-stratified split, achieves a mean Average Precision of 0.984 at Intersection over Union (IoU) of 0.5 (0.956 at IoU 0.5-0.95) on the held-out validation set, with precision 0.988 and recall 0.965. On an unseen held-out test split, mean per-species confidence ranged from 0.96 to 0.99 across the 31 classes, with a 0.17% false-negative rate concentrated in difficult night-time, distant, or occluded images. These metrics are from data from the same pool of sites and cameras as training, so performance at entirely new sites is left to future work. We release the trained weights in ONNX format under a non-commercial licence, with local desktop and real-time camera support, aimed explicitly at ecologists with no machine-learning experience. This release is a deliberate counterweight to the multiple paid for models that have developed over the last decade.

Abstract

Image Classification

Explore similar work

Democratising Camera Trap AI: An Open-Source Model for Detecting UK Mammals

DeepForestVisionV2: Ecology-Driven Taxonomy Expansion for Camera-Trap Monitoring in African Tropical Forests

Camera trap classification with deep learning under ground truth uncertainty