Enemy transporter erector launchers (TEL) carrying intercontinental ballistic missiles (ICBMs) and fighter jets armed with air-to-ground and air-to-air missiles are key targets that friendly forces must quickly locate and shoot down. Countries are spending astronomical amounts of defense budgets to detect air and ground information from the air. Recently, manned and unmanned surveillance and reconnaissance using artificial intelligence (AI) instead of humans has become a global trend.

In Korea, GenGenAI is developing this kind of AI. GenGenAI creates separate data by changing the weather or season in existing images. It can change the look of a road in the rain to a snowing situation, or make it look as if it were taken by a military thermal imaging camera. GenGenAI uses this technology to generate various data for AI training.

GenGenAI shows an AI recognizing ground targets using synthetic data created from real video and the synthetic data itself. /Courtesy of GenGenAI

GenGenAI's technology is being used for an AI pilot under development with Korea Aerospace Industries (KAI). For example, it creates multiple datasets that capture conditions such as day and night and rain on existing transporter erector launcher images so that the system can recognize a TEL in any situation. Going forward, they plan to develop an automatic target recognition function using this data.

The AI pilot is expected to be able to communicate with electro-optical (EO) and infrared (IR) satellites and synthetic aperture radar (SAR). The analysis results from each radar or satellite will also be shared with the AI pilot. This will enable more accurate target recognition.

Cho Ho-jin, CEO of GenGenAI, said, "It will help carry out missions safely and accurately even in extreme environments that are difficult for people to endure."

GenGenAI's synthetic data created from existing images. /Courtesy of GenGenAI

GenGenAI is also cooperating with domestic defense companies to advance weapons systems. It is currently working with Hanwha Systems to study whether synthetic data (artificially generated data) can replace real data in the process of recognizing maritime and aerial targets.

They are also studying ways to use AI created with synthetic data in GOP (General Out Post) perimeter defense and coastal integrated surveillance systems. Cho said, "AI will help partially address the problem of decreasing military strength."

GenGenAI's synthetic data produced by transforming GOP perimeter duty images; it can create synthetic data with various assigned scenarios. /Courtesy of GenGenAI

※ This article has been translated by AI. Share your feedback here.