AI Yoon Seok-yeol, Son Heung-min, and Kim Ju-ha Created Here
Technology Comparable to Global AI Powers... Expertise in Lip Movement

There is an AI company that stands out not only in Korea but also in the AI powerhouses of the United States and China. DeepBrain AI has commercialized AI humans based on deep learning video synthesis technology that closely resemble real people and can engage in real-time conversations. The 'AI Yoon Seok-yeol,' which became a major topic during the 20th presidential election, is a creation of DeepBrain AI. In addition, they have created AI versions resembling celebrities such as soccer player Son Heung-min and anchor Kim Ju-ha.


DeepBrain AI's AI humans are conversational AI solutions that combine deep learning-based video and voice synthesis with natural language processing technology. They provide services across various industries including finance, education, media, entertainment, public sector, healthcare, and commerce. Recently, the financial sector has been actively adopting AI human solutions, with most commercial banks adopting DeepBrain AI's technology. They have also entered the edutech market by creating AI professors for classroom use. Additionally, their solutions have been applied in various fields such as AI show hosts, AI analysts, AI clerks, and AI ambassadors. In February this year, DeepBrain AI was selected in the Enterprise AI avatars category among the top 250 global generative AI startups revealed by CB Insights.

DeepBrain AI AI Human Exhibition Hall [Photo by DeepBrain AI]

DeepBrain AI AI Human Exhibition Hall [Photo by DeepBrain AI]

View original image

DeepBrain AI possesses deep learning video and voice synthesis technology and emotion expression technology that rival those of global companies. They are dedicated to developing AI technology that converses most similarly to humans. Their greatest strength compared to competitors is naturalness. In the industry, the quality of AI humans is often judged by the mouth movements. DeepBrain AI’s mouth shapes, teeth lines, and tongue movements are all natural. Jongho Han, Vice President of DeepBrain AI, said, "The key point that receives positive feedback from customers is quality," adding, "We showcase excellent quality by accumulating know-how through our proprietary algorithms and research institute." About 60-70% of DeepBrain AI’s employees belong to the development organization. The vice president also added, "There are many AI human open sources, and while it is easy to create AI humans by applying them, sometimes the mouth area looks awkward or even lacks lips. DeepBrain AI ensures that customers do not feel discomfort when the technology is applied."


Based on this technological capability, they are targeting the global market. They have established local subsidiaries in the United States and China. To target the US and European markets, CEO Jaeng Se-young has been stationed at the US subsidiary for the past year, leading the efforts. Recently, the number of cases adopting DeepBrain AI’s solutions in the global market has been increasing. In the US, they created an AI human of comedian Howie Mandel, the longest-serving judge on the famous audition program "America's Got Talent." The AI announcer for China’s state broadcaster CCTV was also produced using DeepBrain AI technology. They have also signed contracts with several other global companies. The vice president confidently stated, "The quality of AI humans is the best in Korea and can be considered the best even compared to global companies."

AI Human [Photo by DeepBrain AI]

AI Human [Photo by DeepBrain AI]

View original image

In line with the recent ChatGPT craze, they are also preparing to combine generative AI technology with their AI humans. Connecting generative AI technology with AI humans enables real-time conversations that feel like talking to actual humans. The vice president said, "If generative AI like ChatGPT becomes established in each company, the AI human market will also grow," adding, "Currently, it takes more than 4 seconds to receive an answer when asking ChatGPT, and with DeepBrain AI’s real-time synthesis technology, it takes 6 seconds. However, if individual companies create their own language models to provide answers within 1 second and complete AI human speech within 2-3 seconds, the potential for service expansion is huge." They are also preparing for the full-scale advent of the metaverse era. The vice president said, "After the emergence of ChatGPT, the enthusiasm for the metaverse cooled down, but I believe it was because it was based on characters rather than real images, which had limitations," adding, "To expand the metaverse, although it is a virtual space, it must be based on real images. We will create 3D virtual humans so that AI humans can naturally play in the metaverse space. We are also conducting technology verification (POC) with large corporations."



In the long term, DeepBrain AI plans to establish itself as a global leading company in the AI human market. The vice president stated, "We will embed the formula 'AI Human = DeepBrain AI' in the global market." He continued, "The company’s goal is to create the future with innovative and human-centered AI technology," adding, "The ultimate goal is to create AI human services that provide great help to humans."


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

© The Asia Business Daily(www.asiae.co.kr). All rights reserved.

Today’s Briefing