Nexon Partners with Google Cloud to Enhance Harmful Image Detection Using AI

Nexon announced on the 26th that it has enhanced the performance of its harmful image detection artificial intelligence (AI) model in collaboration with Google Cloud.

Nexon actively utilizes AI technology to provide an optimized gaming environment for 1.9 billion users worldwide. Among these, the harmful image detection system is a technology that detects inappropriate elements such as obscenity, hate, violence, and discrimination in content shared with users by using a self-developed AI model.

Nexon CI [Image source=Nexon]

Nexon's harmful image detection system independently developed AI models tailored to the unique characteristics of each game, such as different selection criteria and character designs. To improve and advance the detection model's performance in line with rapidly evolving AI technology, Nexon collaborated with engineers from Google. This collaboration is the first case among domestic game companies to introduce AI technology for harmful image detection in partnership with a global AI company.

Nexon not only used images collected independently but also constructed an optimal dataset specialized for animation and games to train the AI. Through collaboration with Google Cloud, tens of terabytes of large-scale data training were conducted quickly and effectively. Nexon plans to strengthen its collaboration with Google to further develop its AI technology capabilities.

Jung Chang-wan, Head of Advanced Development at Nexon Intelligence Labs, said, "Through the collaboration project with Google Cloud, we gained overall know-how and confidence in AI-related advanced technologies and confirmed the possibility of actively utilizing AI technology in more areas." He added, "We will continue to do our best to create an environment that provides a better gaming experience by utilizing various AI models."

This content was produced with the assistance of AI translation services.