Look, let's be honest: Netmarble has managed to turn quite a few heads by clinching two first-place titles at the 'Object Recognition Challenge' during the 2025 International Conference on Computer Vision (ICCV), as announced on the 24th. The ICCV, a biennial gathering of the brightest minds in computer vision, held its latest event in Hawaii from the 19th to the 23rd.

The 'Object Recognition Challenge' is essentially a high-stakes arena where AI models compete to see who can best estimate the 2D and 3D properties of objects within images. Netmarble’s internal AI tech proved its worth, securing top spots in both the object segmentation and lightweight recognition categories.

It wasn't just about winning prizes, though. The team also presented a paper titled 'MUSE (Model-based Uncertainty-aware Similarity Estimation)' at the conference’s workshop. Their research highlights a clever way for AI to identify and segment objects without the need for additional, tedious training sessions.

For those keeping track, Netmarble has been pouring resources into AI research—covering everything from computer vision to natural language processing and generative AI—since 2014, all with the goal of streamlining their game development pipeline and sharpening their competitive edge. Adieu.