The work of a student at the University of Alberta (Edmonton, Canada) demonstrated how artificial intelligence technologies can make it easier to study birds, the school’s newsletter reports.
Priscilla Adebanji, a computer science student, spent the summer experimenting with AI to improve the analysis of tracking videos of red-winged blackbirds and their nests.
Its solution, if fully implemented, will save the hours required for manual identification of specific birds, confirmed Professor Ivana Schopf, who studies the effect of parasite infection on the behavior of birds.
Current methods involve recognizing birds and their behavior from their calls. This requires hours of reviewing records by two employees (for consistency) who require a certain level of knowledge and experience, she noted.
While existing software can track the movements of animals like mice in the laboratory, it is more difficult to do so in nature due to suboptimal conditions, Schopf said. Another problem is the quality of the video, since the nests are well hidden in the marsh vegetation.
Schopf approached Adebanji’s faculty member, Professor Thibault Lutelier, looking for a way to automatically detect birds without manual reviews and provided recordings from two field seasons—a total of 30 hours of video.
“We felt there were many AI applications that could help, although we had no idea what to expect. We needed to figure out what type of machine learning to use. We did a lot of preparatory work and research,” said Lutelier.
Adebanji had to overcome various problems, including false readings created by existing AI models.
“Sometimes they would mistake a bird for an airplane and misidentify things like shadows as bears in the background,” the student explained.
Using computer vision tools and motion detection algorithms to analyze video, she improved tracking quality enough to count birds, separate them from other objects, and identify them.
By the end of the summer, the software developed by Adebanji was able to determine the exact time when birds fly into and leave their nests. According to her, automatic detection has reduced the time required for this from eight hours to a couple of minutes.
While the software still needs work, Schopf hopes to eventually use it to eliminate the time-consuming process of manually collecting data for his research. In her opinion, the solution has the potential for wider application in the study of animals.
Recall that in May, the famous critic of Bitcoin, billionaire Warren Buffett compared AI to an atomic bomb and expressed alarm about the rapid development of technology.
Found an error in the text? Select it and press CTRL+ENTER
ForkLog newsletters: keep your finger on the pulse of the Bitcoin industry!