DGIST Faculty and Researchers Selected as Final Presenters at Top AI Conference CIKM 2025

October 14, 2025

Unofficial English summary. Original coverage: read in Korean.

A team led by Professor Ji-woong Choi of DGIST’s Department of Electrical Engineering and Computer Science, working with TORIS, founded by DGIST Senior Researcher Dae-gun Oh, was selected to present at the Industry Day Talks of CIKM 2025, held at COEX in Seoul from November 10 to 14, Electronic Times reported.

Their AI anti-drone technology extends small-drone identification range from 2–3 km to 8 km, which the team described as a world first. Earlier systems lacked radar and infrared cameras able to track targets beyond 8 km, so the data needed to train long-range AI did not exist.

The team trained a 3D convolutional neural network on long-range video from TORIS’s radar–IR system. The model learned the motion differences between drones, which fly in regular patterns, and birds.

The system identified small drones at 8 km, remained stable on low-resolution footage, identified drone models not seen in training with high accuracy, and sharply reduced false alarms from birds.

The first author, PhD student Woochul Jin, worked at TORIS through an industry–academia program on radar and IR sensors. Professor Choi and Senior Researcher Sang-Chul Lee were co-corresponding authors, and Dae-gun Oh took part in the research.

Oh said the result shows that DGIST’s AI combined with TORIS’s world-class hardware can solve hard problems in defense and security. Choi called it a representative success of industry–academia–research cooperation.

Other outlets added details: Hankook Ilbo quoted Oh as saying his final goal is to prove TORIS’s anti-drone technology is the best in the global defense market, and Digital Daily pointed to uses such as protecting airports, ports and nuclear power plants.

Also Reported By