DeepFauna Sub-Saharan Africa: ONNX Model Weights for Sub-Saharan African Species Classification

Description:

DeepFauna Sub-Saharan Africa is a species classifier for camera trap imagery covering Sub-Saharan African wildlife, developed as part of the Trap Tracker platform (traptracker.co.uk). It assigns one of 36 class labels to detected animal/person/vehicle crops, working downstream of the companion Sub-Saharan Africa Wildlife Detector. Weights are exported in ONNX format for cross-platform inference without requiring the original training framework.

Supported classes (36):

Cheetah (Acinonyx jubatus jubatus), Car, Blue wildebeest (Connochaetes taurinus), Plains zebra (Equus quagga), Baboon (Papio sp.), Giraffe (Giraffa camelopardalis), African elephant (Loxodonta africana), Lion (Panthera leo), Person, Rhino, African buffalo (Syncerus caffer), Eland (Tragelaphus oryx), Spotted hyena (Crocuta crocuta), Impala (Aepyceros melampus), Black-backed jackal (Canis mesomelas), Sable antelope (Hippotragus equinus), Rabbit (Oryctolagus cuniculus), Warthog (Phacochoerus africanus), Chimpanzee (Pan troglodytes), Giant pangolin (Smutsia gigantea), Crested porcupine (Hystrix cristata), Aardvark (Orycteropus afer), Hippopotamus (Hippopotamus amphibius), Gemsbok (Oryx gazella), Ostrich (Struthio camelus), Hartebeest (Alcelaphus buselaphus), Waterbuck (Kobus ellipsiprymnus), Gorilla (Gorilla sp.), Bongo (Tragelaphus eurycerus), Kob (Kobus kob), Helmeted guineafowl (Numida meleagris), Common duiker (Sylvicapra grimmia), Genet (Genetta sp.), Giant pouched rat (Cricetomys sp.), Vervet monkey (Chlorocebus tantalus), Leopard (Panthera pardus).

Training data:

The model was fine-tuned on 56,131 labelled animal/person/vehicle crops, split into training (45,059), validation (5,491), and test (5,581) sets. Data was collected through Sub-Saharan African conservation deployments, spanning daylight RGB and infrared night-time imagery, motion blur, partial occlusion, and multi-animal frames.

Raw training images are not released, as many were provided under partner-specific data agreements that pre-date this open model release.

Performance:

On a held-out test set of 7,338 crops, DeepFauna Sub-Saharan Africa achieves a macro-averaged F1 score of 0.9922 (overall accuracy 0.992) across all 36 classes. Performance is strong and even across species; the lowest-scoring classes are Gorilla (F1 0.968), Rhino (F1 0.979), and Buffalo (F1 0.981) — Rhino's lower recall (0.963) likely reflects visual overlap with other large grey-bodied megafauna at distance or poor light. All other classes score F1 ≥ 0.98, with 11 classes achieving a perfect F1 of 1.000.

Intended use:

For conservation practitioners, ecologists, and citizen science projects working across Sub-Saharan Africa to automate species identification in camera trap and field imagery. Released freely for non-commercial conservation, research, teaching, and citizen science use, particularly for under-resourced organisations lacking compute or ML expertise to train from scratch.

Related publication: No accompanying paper yet. This release is documented via this Zenodo record and its included test report and dataset manifest.

License: CC BY-NC 4.0 (Creative Commons Attribution-NonCommercial 4.0 International)

Citation: Please cite this Zenodo DOI when using this model in published work. Commercial use requires separate permission.