Artificial intelligence helps with wildlife conservation efforts

Just as the footprints of crime scenes can be important evidence, the footprints of wildlife are equally valuable for wildlife conservationists.

Each species - and every animal within that species - has its unique footprint. Each footprint and its underlying matrix tells a story about animal activity and the environment. The information collected from footprints helps conservationists track the behavior of wildlife and monitor endangered species. This information also helps to protect biodiversity on Earth. To train the algorithm, researchers lured tigers and other wild animals on a prepared piece of land. They took multiple images of each foot - four feet per animal is equivalent to four times the amount of data that can be collected.

The initial footprint recognition now includes other biometric features, including fur color, fur type, and vocalizations. WildTrack can monitor migration patterns, mating activity, foraging and hunting behavior, as well as social groups. The different substrates beneath the footprints indicate humidity levels, weather conditions, and gait changes, the latter of which can indicate walking speed or injury signals.

The adjustment of machine learning algorithms will also support research in other fields. Artificial intelligence based solutions can be used for statistics and monitoring of endangered animals, as well as preventing poaching. Future uses may include tracking sea turtles based on their unique shell shape, identifying the types and health status of animals and plants, or detecting and monitoring traces of animal or human activities.