AI Experts Develop Big Data Approach for Wildlife Preservation

A group of artificial intelligence (AI) and animal ecology experts at Ecole Polytechnique Fédérale de Lausanne have developed a new big data approach to enhance research on wildlife species and improve wildlife preservation. 

The new study was published in Nature Communications. 

Collecting Data on Wildlife

The field of animal ecology now relies on big data and the Internet of Things, with massive amounts of data being collected on wildlife populations through technology like satellites, drones, and automatic cameras. These new technologies result in faster research developments while also minimizing disruption in natural habitats. 

Many AI programs are used to analyze large datasets, but they are often general and not precise enough to observe the behavior and appearance of wild animals. 

The team of scientists developed a new approach to get around this, and they did so by combining advances in computer vision with the expertise of ecologists. 

Leveraging Expertise of Ecologists

Ecologists currently use AI and computer vision to extract key features from images, videos and other visual forms of data, which enables them to carry out tasks like classifying wildlife species and counting individual animals. However, generic programs that are often used to process this data are limited in their ability to leverage existing knowledge on animals. They are also difficult to customize and are prone to ethical issues related to sensitive data. 

Prof. Devis Tuia is the head of EPFL’s Environmental Computational Science and Earth Observation Laboratory and lead author of the study. 

“We wanted to get more researchers interested in this topic and pool their efforts so as to move forward in this emerging field. AI can serve as a key catalyst in wildlife research and environmental protection more broadly,” says Prof. Tuia.

In order to reduce the margin of error of an AI program that is trained to recognize a specific species, computer scientists would need to be able to leverage the knowledge of animal ecologists. 

Prof. Mackenzie Mathis is the head of EPFL’s Bertarelli Foundation Chair of Integrative Neuroscience and co-author of the study. 

“Here is where the merger of ecology and machine learning is key: the field biologist has immense domain knowledge about animals being studied, and us as machine learning researchers' job is to work with them to build tools to find a solution,” she said. 

This is not the first time that Tuia and the team of researchers has addressed this issue. The team previously developed a program to recognize animal species based on drone images, while Mathis and her team have developed an open-source software package to help scientists estimate and track animal poses. 

As for the new work, the team hopes it can capture a wider audience.

“A community is steadily taking shape,” says Tuia. “So far we've used word of mouth to build up an initial network. We first started two years ago with the people who are now the article's other lead authors: Benjamin Kellenberger, also at EPFL; Sara Beery at Caltech in the US; and Blair Costelloe at the Max Planck Institute in Germany.”