A machine learning system that can isolate potential illegal wildife trade products for sale on online marketplaces. Current training data is on ivory, pangolin, wild cat teeth and claws, elephant hair and skin. The system is housed at WWF Singapore and is part of the Cyber Spotters initiative. Data extrapolated from this system gets forwarded to law enforcement (if deemed serious), to e-commerce companies (for their records) and kept internally to be used as references for potential future action i.e. demand reduction initiatives, digital deterrent.

Pilot machine learning solutions in the process of being developed to assess and predict locations where known drivers of zoonotic disease emergence, spillover and outbreaks are likely to occur. This solution will utilise open source environmental, epidemiology and demographic data including deforestation data, disease outbreaks reports, wildife market location evaluations and regional infrastructure developments. The ML algorithms will be tailored to extract features within defined geographical coordinates, utilizing spatial analytics to assess proximity to relevant geographical entities. The algorithm will then seek to identify the importance of geographic, demographic, environmental or other factors in predicting disease hotspots.

AI Guardian 2.0 uses a customized model created by PaddleX, a development tool for the PaddlePaddle platform. The tool uses internet technology to enable the comprehensive management of illegal wildlife trade on the Internet and to make wildlife protection more efficient. The tool adopts the latest large vision model (LVM) and was trained jointly by the International Fund for Animal Welfare (IFAW) and Baidu PaddlePaddle through semi-supervised learning. The model will be upgraded regularly to incorporate the latest technology development and will expand the number and variety of species covered to keep up with latest trends in wildlife cybercrime.

A complete, self-contained, multi-illumination, field-deployable, open-source platform for field imaging and identification of forest products at the macroscopic scale. The XyloTron platform integrates an imaging system build with off-the-shelf components, flexible illumination options with visible and UV light sources, software for camera control, and deep learning models for identification.

Intelligently processes images from camera traps in the field and immediately flags to rangers the presence of an elephant or human, allowing those protecting wildlife to identify poaching threats in real time and respond more immediately. The tool is not actively maintained or available for use at this time.

Using artificial intelligence and the power of big data to provide scientists an unequaled view into the habits and habitats of wildlife, data that is critical for crafting smart conservation policies. It takes “animal selfies” through motion-detector cameras — known as camera traps — to snap thousands of photos a day of animals rarely seen by human eyes. As the largest camera-trap database in the world, Wildlife Insights has the potential to transform wildlife conservation by providing reliable, frequent and up-to-date information on myriad species that are largely invisible to science and conservation practitioners.

Maritime risk analytics platform used to identify illicit activity through inconsistent or “risky” behavior.

An extension of the SCAT tools. It aims to determine the geographic origin(s) of large elephant ivory seizures using a Voronoi tessellation method that utilizes genetic similarities across tusks to simultaneously infer the origin of multiple samples that could have one or more origin(s).

A machine learning-based dashboard that captures shipping invoices (electronic or paper) and conducts analytics on cargo shipment data (ex. price, weight, country of origin, taxonomic identity) for inspectors to easily identify questionable shipments. It is being tested with Canadian government agencies. It was designed to automatically analyse shipment paperwork and identify questionable shipments. Once anomalies have been identified and shipments are being inspected, officers will be supported in their decision making with an integrated species identification application.

AI-driven detection algorithm for computed tomography scanning systems using bespoke Microsoft AI to curb the trade. Currently, there are no integrated, multi-species illegal wildlife algorithms on the market. The project can be analysed to build a comprehensive and detailed picture of illegal wildlife trafficking across the world. It has been developed in collaboration with the United Kingdom Border Force’s Convention on International Trade in Endangered Species (CITES) team, London Heathrow Airport, as well as the Duke of Cambridge’s Royal Foundation.