Artificial Intelligence Puts Focus on the Life of Insects

% of readers think this story is Fact. Add your two cents.

“With the help of advanced camera technology, we can now collect millions of photos at our field sites. When we, at the same time, teach the computer to tell the different species apart, the computer can quickly identify the different species in the images and count how many it found of each of them. It is a game-changer compared to having a person with binoculars in the field or in front of the microscope in the lab who manually identifies and counts the animals”, explains senior scientist Toke T. Høye from Department of Bioscience and Arctic Research Centre at Aarhus University, who headed the new study. The international team behind the study included biologists, statisticians, and mechanical, electrical and software engineers.

The methods described in the paper go by the umbrella term ‘deep learning’ and are forms of artificial intelligence mostly used in other areas of research such as in the development of driverless cars. But now the researchers have demonstrated how the technology can be an alternative to the laborious task of manually observing insects in their natural environment as well as the tasks of sorting and identifying insect samples.

Automatic moth trap with moths detected by computer vision

Cedit: Toke T. Høye)

“We can use the deep learning to find the needle in the hay stack so to speak – the specimen of a rare or undescribed species among all the specimens of widespread and common species. In the future, all the trivial work can be done by the computer and we can focus on the most demanding tasks, such as describing new species, which until now was unknown to the computer, and to interpret the wealth of new results we will have”, explains Toke T. Høye.

And there is indeed many tasks ahead, when it comes to research on insects and other invertebrates, called entomology. One thing is the lack of good databases to compare unknown species to those which have already been described, but also because a proportionally larger share of researchers concentrate on well-known species like birds and mammals. With deep learning, the researchers expect to be able to rapidly advance knowledge about insects considerably.

Long time series are necessary

To understand how insect populations change through time, observations need to be made in the same place and in the same way over a long time. It is necessary with long time series of data.

Deep Learning 

Credit: Kim Bjerge

Some species become more numerous and others more rare, but to understand the mechanisms that causes these changes, it is critical that the same observations are made year after year.

An easy method is to mount cameras in the same location and take pictures of the same local area. For instance, cameras can take a picture every minute. This will give piles of data, which over the years can inform about how insects respond to warmer climates, or to the changes caused by land management. Such data can become an important tool to ensure a proper balance between human use and protection of natural resources.

“There are still challenges ahead before these new methods can become widely available, but our study points to a number of results from other research disciplines, which can help solve the challenges for entomology. Here, a close interdisciplinary collaboration among biologists and engineers is critical”, says Toke T. Høye.
Contacts and sources: 

Publication: Deep learning and computer vision will transform entomology.
Toke T. Høye, Johanna Ärje, Kim Bjerge, Oskar L. P. Hansen, Alexandros Iosifidis, Florian Leese, Hjalte M. R. Mann, Kristian Meissner, Claus Melvad, and Jenni Raitoharju. PNAS, 2021 DOI: 10.1073/pnas.2002545117