The National Park Service is set to implement artificial intelligence (AI) to analyze wildlife videos, significantly reducing the time required for analysis. The time to analyze 100,000 videos is expected to decrease from 208 days to just 8 hours.
On October 13, the Ministry of Environment's National Park Service announced the recent development of 'Bandal Lens AI,' which automatically identifies species of domestic wildlife and plans to apply it in the field.
Approximately 4.7 million videos are collected annually from over 1,300 unmanned sensor cameras installed in national parks across the country. Previously, human analysts reviewed the footage to distinguish between different animal species.
Bandal Lens AI can automatically identify 17 mammal species captured by the cameras, including Asiatic black bears, mountain goats, foxes, otters, martens, and wildcats. Dogs and cats are also included in the identification process.
According to the National Park Service, analyzing 100,000 videos with existing personnel would take 208 days, but with this new system, it will be reduced to just 8 hours. The goal is to automate repetitive video classification tasks, thereby enhancing the expertise in species and habitat conservation efforts.
The agency plans to further enhance the system for user-friendliness and will make the AI model available on Hugging Face and GitHub. This will support domestic and international research institutions and organizations that require mammal video analysis.
Joo Dae-young, the director of the National Park Service, stated, "The introduction of this system brings us closer to automating the management of natural resources and habitats in national parks. We will further strengthen the expertise in species and habitat conservation in light of the increased efficiency of our work."
* This article has been translated by AI.
Copyright ⓒ Aju Press All rights reserved.