Naver Adopts 8 Papers at Global Conferences... "Securing AI Leadership"

Naver announced on the 13th that eight of its papers were accepted at NeurIPS (Conference on Neural Information Processing Systems), the world's most prestigious artificial intelligence (AI) conference. This is the highest number ever.

NeurIPS is a conference where mid- to long-term research results in the field of machine learning are mainly presented, sharing technological achievements from a long-term perspective. This year, it will be held from December 6 to 14.

Naver will present research outcomes proposing methods to innovatively improve the training of AI models. Unlike existing deep learning trends, it introduces research that appropriately imposes constraints on models. It suggests ways to overcome the limitations of existing deep learning models, which solved problems differently from humans but suffered from reduced reliability.

In addition, a paper proposing a new optimization technique to address the problem that currently used AI models cannot flexibly respond to situations different from training was accepted at the conference. Another paper proposing a learning method that can add new object categories while preserving the memory of already known object information will also be presented at the conference.

The company explained that these achievements are the result of steady investment in research and development (R&D) from a long-term perspective. Naver invests about 25% of its sales in R&D, with a large portion dedicated to the AI field.

Ha Jung-woo, head of Naver AI Lab, said, "The core competitiveness of an AI technology platform company lies in strengthening its own research capabilities," adding, "We will continue long-term investments in AI technology, maintain various research collaborations, and secure global AI leadership."

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