Open-source AI model for UK camera trap species detection
The TrapTracker UK Mammals Model is an open-source object detection model designed for UK camera trap images. It has been developed to help conservation organisations, ecologists, universities, citizen science groups, and statutory agencies process large volumes of camera trap data without needing commercial platforms or machine-learning expertise.
The model detects and classifies 31 UK-relevant classes, including common mammals, selected bird species, domestic animals, people, vehicles, and operational utility classes such as calibration poles.
It is released as part of our wider aim to make conservation AI more accessible, transparent, and useful to the organisations that need it most.
Model overview
The model is a YOLO26x object detector trained on a curated UK camera trap dataset built from nearly a decade of Conservation AI and TrapTracker deployments.
It was developed specifically for UK conservation settings, rather than being adapted from a general wildlife model or a model trained primarily on non-UK fauna. This matters because camera trap AI performance often depends heavily on the species, landscapes, camera types, lighting conditions, and deployment contexts represented in the training data.
The model is intended to provide a strong, practical baseline for UK camera trap analysis, particularly for users who need local, offline, inspectable, and non-commercial tools.
What the model detects
The model covers 31 classes:
- European hedgehog
- Grey squirrel
- Red squirrel
- Roe deer
- Red deer
- Red fox
- European badger
- Person
- Domestic goat
- European rabbit
- Wood pigeon
- Common pheasant
- Domestic sheep
- House sparrow
- Fallow deer
- Domestic cattle
- Domestic cat
- Pine marten
- Domestic dog
- Calibration pole
- Northern goshawk
- Common buzzard
- Capercaillie cock
- Capercaillie hen
- Eurasian curlew
- Curlew chick
- Car
- Domestic horse
- Carrion crow
- Wild boar
- Reeves’ muntjac
Please note that Car is included as a reserved class in the taxonomy but is not currently populated with labelled training instances, so the model should not be expected to predict this class reliably in practice.
Training data
The model was trained on labelled instances drawn from UK camera trap images collected through conservation deployments over nearly a decade.
The dataset includes a wide range of practical camera trap conditions, including:
- daylight RGB images
- infrared night-time images
- motion blur
- partial occlusion
- multi-animal frames
- juveniles and adults
- manufacturer overlays and timestamps
- varied UK habitats and deployment settings
The raw training images are not released, because many were provided under partner-specific data agreements that pre-date this open model release.
Training setup
The model was trained using:
- Architecture: YOLO26x
- Input resolution: 640 × 640
- Training hardware: 8 × NVIDIA RTX A6000 GPUs
- Training schedule: 120 epochs
- Global batch size: 256
- Framework: Ultralytics
- Export format: ONNX
- Training time: approximately 8 hours 16 minutes
The model was trained with deterministic settings and a fixed random seed to support reproducibility.
Performance
On the held-out validation split, the model achieved:
| Metric | Value |
|---|---|
| Precision | 0.988 |
| Recall | 0.965 |
| mAP@0.5 | 0.984 |
| mAP@0.5:0.95 | 0.956 |
| Optimal F1 confidence | 0.620 |
| F1 score at optimal confidence | 0.98 |
On an additional class-stratified held-out test split, the model produced 4,820 confident detections from 4,838 test instances, with only 3 images flagged for manual review.
These results indicate strong in-distribution performance across the 31-class UK camera trap detection task.
Important limitations
The headline results should be interpreted carefully.
The validation and test results are in-distribution, meaning they are drawn from the same overall data pool as the training data. The model has not yet been fully benchmarked on a formal site-disjoint or temporally-disjoint out-of-distribution test set.
Performance may therefore be lower when the model is used on:
- new camera trap locations
- unusual habitats
- different camera models
- poor-quality images
- rare behaviours or unusual poses
- species or classes not represented in the model taxonomy
- regions outside the UK
Users should treat the model as a strong practical baseline, not as a replacement for ecological expertise or final human review.
Intended use
This model is intended for non-commercial conservation and ecological monitoring, including:
- UK camera trap species detection
- biodiversity monitoring
- ecological research
- conservation project workflows
- university and student research
- citizen science projects
- local offline image processing
- pre-filtering large image collections before manual review
The model can be used through TrapTracker tools or integrated into other non-commercial workflows using the ONNX weights.
Not intended for
The model is not intended for:
- commercial resale or paid inference services
- use as a general global wildlife model
- legally sensitive identification tasks
- surveillance or monitoring of people
- replacing expert ecological validation
- making conservation decisions without human review
- detecting species not included in the model classes
Where the results are used in scientific, regulatory, or conservation decision-making, users should include appropriate manual checking and validation.
Why we are releasing it
Camera trap AI has become increasingly important for biodiversity monitoring, but access to high-performing species recognition tools is often limited by commercial platforms, proprietary weights, pay-per-image models, or cloud-based inference.
This release is intended as a practical counterweight to that trend.
The aim is simple: a conservation organisation with limited funding, no machine-learning team, and no GPU should still be able to process camera trap images locally and obtain useful species-level detections.
The model is free for non-commercial use, runs offline, and is designed for ecologists and conservation practitioners rather than machine-learning specialists.
How to cite
If you use the model in research, reports, publications, or conservation outputs, please cite:
Democratising Conservation AI: An Open-Source TrapTracker Model for UK Mammals.
A formal citation will be added here once the paper is available online.
Licence
The model weights are released for non-commercial use.
You may use the model for conservation, research, teaching, ecological monitoring, citizen science, and non-commercial biodiversity work.
You may not resell the model, use it to provide a paid commercial inference service, or incorporate it into a commercial product without prior permission.
For commercial licensing enquiries, please contact:
Download
By downloading the model, you agree to use it only for non-commercial purposes and to acknowledge the model appropriately in any resulting publications, reports, or public outputs.
Updates
The Trap Tracker UK Mammals Model will continue to be developed as new data become available and as the underlying model architecture improves. Future releases will be made available periodically, with clear versioning so that users can track changes, improvements, and compatibility across model updates.