Using AI, IPB University Research Team Successfully Assesses the Beauty and Health of Coral Reefs
Marine experts from IPB University have successfully utilized artificial intelligence (AI) to quantify the beauty of coral reefs with near-maximum accuracy. The success of the IPB University research team may well become the standard for researchers worldwide.
Previously, the Mars Coral Reef Restoration Project team conducted coral reef restoration efforts in South Sulawesi. These efforts were monitored by a joint team, including researchers from IPB University.
This research is a collaboration between the IPB University research team and international researchers from Lancaster University, the University of Montpellier, and the University of Texas.
There are 18 coral reef sampling sites located around Bontosua Island, South Sulawesi. Researchers selected coral reef sites from an area spanning 0,79 km².
The research program is led by Cut Aja Gita Alisa, a graduate student in the Marine Science Program at IPB University. It involves Dr Beginer Subhan, Dr Tries B. Razak, and Prof Neviaty P Zamani, as well as other graduate students, including Rindah Talitha Vida.
The research findings have been published in the reputable journal, Scientific Reports and featured on Nature.com under the title “Benthic communities on restored coral reefs confer equivalent aesthetic value to healthy reefs.”
“We used AI in this research. AI helped us make decisions to quantify the beauty of restored coral reefs,” explained Indonesian coral reef restoration expert Dr Beginer Subhan recently.
“We hope that the results of this research can become a standard for evaluating the success of coral reef restoration activities in Indonesia, and even worldwide,” he added.
The foreign researchers involved include Tim Lamont, Nicolas Mouquet, Nicholas Graham, Christopher Hemingson, and David Mouillot. The entire monitoring team of the Mars Coral Restoration Project, a group of Indonesian marine researchers based in Makassar, also contributed.
Regarding AI involvement, Gita Alisa explained that they also used underwater photography techniques with an Olympus TG-5 camera, which was then evaluated by AI. “The result is R-squared equals 0,95. In many studies worldwide, an R-squared value approaching 1 falls into the category of high prediction accuracy or very high accuracy,” Gita Alisa elaborated.
Thus, AI successfully quantified that the restoration results were highly similar to natural coral reefs. This finding confirms that proper restoration can restore the beauty of coral ecosystems to their natural condition.
Monitoring Method
Indonesia is the country with the most coral reef restoration programs among nations worldwide. The restoration method studied in this research is the “Reef Star” method initiated by the Mars Coral Restoration Project. This method involves restoring coral reef fragments, enabling coral coverage in the sea to increase from 10% to 60%.
Before the AI assessment, researchers randomly distributed 883 coral reef photos with various conditions (healthy, restored, damaged) to 3.348 respondents worldwide via online connections.
These thousands of respondents came from France, Indonesia, the United Kingdom, the United States, and Australia. They represented diverse backgrounds: academics, environmental advocates, non-profit organizations (NGOs), and the scientific community.
The results of this online survey were consistent, with respondents selecting the most beautiful coral reef photos featuring vibrant colors and complex coral structures. This underscores the importance of visual diversity in attracting public interest in coral reefs.
AI Utilization
By combining restoration methods with AI, researchers at IPB University have successfully predicted the aesthetic value of coral reefs in restoration areas, which is equivalent to that of healthy natural reefs.
According to Dr Beginer, in this era of social media and nearly unlimited information access, the use of AI must be done with a focus on maximizing positive benefits. That is why, he said, IPB University is utilizing AI in this research to promote the use of AI technology in marine conservation.
“Going forward, IPB University will utilize AI to monitor ecosystem recovery efficiently and objectively, supporting marine protection efforts in Indonesia and around the world,” Dr Beginer stated, highlighting the university’s commitment. (IAAS/FMT)
