Face Recognition Venture Finds Success with Big Companies
Naver invested in a new face recognition venture as soon as it was established.
At the same time when a company was founded, there is a startup that received 1.5 billion won in investment from Naver. Naver valued its 30 percent stake at 1.5 billion won. As soon as the company was established with a capital of 70 million won, it was recognized as a corporate value of 5 billion won. This is the story of Alchera, a group of engineers who dreamed of becoming artificial intelligence and augmented reality experts.
Alchera is a new technology startup that was established in June last year and is still less than a year old. Kim Jung-bae, a former CEO of the Samsung Advanced Institute of Technology, Hwang Young-kyu, vice president of the company, and Oh Young-taek, chief technology officer, co-founded the company. He chose Alchera, which means “Dream Time” in Aboriginal language, as his mission to achieve the meaning of “a dream with AI.”
“We want Alchera to become the most attractive company that researchers want to work within AI and AR fields, and thus open an era of dreams for mankind,” said CEO Kim Jung-bae. “The most important thing in technology development is to research and discuss new technologies with a great engineer who is highly skilled and creative.”
Dreams held by graduates of Samsung Advanced Institute of Technology for 6 years
Kim earned his master’s and doctorate degrees from KAIST through computer vision and machine learning. Later, he served 13 years at the Samsung Advanced Institute of Technology. He was recognized as an expert in facial recognition technology. CEO Kim is credited with the face recognition technology that was applied to Galaxy smartphones last year. Hwang and Oh also worked with Kim for nine and five years, respectively, at the Samsung Advanced Institute of Technology.
Three co-founders specialized in computer vision and computer graphics developed the world’s first fully automatic medical fusion imaging technology. They made that technology when they were working together at Samsung Advanced Institute of Technology. Kim said. “This technology won the Samsung thesis gold prize and was commercialized under the name of S-Fusion through the medical device division.”
It took six years for the three to create the first company jointly. Kim made the proposal with his own ideas over the weekend after investing in the activities with Hwang. Based on this, it has been making efforts for four years by holding presentations and making patents.
As Hwang left Samsung and moved to SK Telecom’s Future Technology Institute, all of the plans he wanted to do business with seemed to be falling apart. After one and a half years of vacancies, Vice President Hwang in Snow’s subsidiary, Naver, started a 3D-related technical consultation and took the opportunity. With the technology of co-founders gaining recognition, Alchera will also receive investments from Naver at the same time as the company is set up.
Alchera includes 12 research workers, including six Ph.D., three master’s and three bachelor’s degrees, out of 14 positions. Aside from the project meeting, they are holding a technical seminar, which is a form of two weekly introductions to the new technology and discussions. To this end, they read and discuss papers and share ideas, which is a valuable time for the researchers.
“Many countries, including the United States, Japan, Britain, Israel, and China, are supporting artificial intelligence research and development at a national level, and global companies are investing a lot,” Kim said. “I want to lead the world in this field by creating attractive companies that good researchers want to work within the field of artificial intelligence that is just in motion.”
*New technology is introduced through various channels other than Snow application
Alchera has developed an AR-based 3D face recognition engine. It is a technology that can recognize characters in videos and add 3D virtual objects by using Deep Learning. Others are developing facial recognition mostly for authentication purposes. However, the engine develops and uses effects filters in video apps. In the future, the company plans to expand its scope to body recognition and object recognition to showcase the product in a wider variety of applications and channels.
Alchera’s 3D face AR engine has been adopted by Snow, Asia’s No. 1 player in-camera apps. In addition, with sensor-based 3D AR technologies being mounted on snow, Snow played an important role in expanding its service from 2D AR to 3D AR. After the Snow partnership, other companies are also interested in Alchera’s technical skills and will be contacted.
“Naver’s investment in the early stages of its start-up is right for snow cameras to be equipped with 3D AR engines. However, Alchera’s business is not limited to Snow,” Kim said. “Our new technology will be introduced through many channels in addition to Naver and Snow soon.” “We are also thinking of attracting additional investments to develop advanced artificial intelligence engines by recruiting more researchers.”
In fact, Alchera deals with both the image recognition field of extracting information from input images and the augmented reality field of showing three-dimensional information back to people. Alchera researchers have a lot of experience in general object recognition and medical imaging as well as face recognition. Hwang studied facial recognition technologies such as 3D facial expression tracking and facial cloning, and medical fusion imaging, while Oh CTO has worked to combine computer graphics technologies into medical images.
Alchera plans to expand its scope to recognize hand gestures, recognize the movement of body joints, separate body, and background, and restore 3D to the body. There are also many new projects such as augmented reality without markers.
“Alchera has its own deep learning framework and AR engine that can operate in real-time on smartphones,” said CEO Kim. “It is our biggest competitive position because it’s hard to find a company that has all the necessary processes in AI and AR. And We have the learning data production team to deep learning algorithm research team to mobile optimization team.”