#Power of Businesses l 2021-04-26
Now it’s time to take a look at a Korean business bringing about changes in the global economy with some new ideas.
Today, we’re going to introduce you to Neurocle, a developer of deep learning solutions for non-professionals. Let’s hear from company CEO Lee Hong-suk (이홍석).
Neurocle aims to enable anyone to use artificial intelligence or AI technology. We focus on the field of deep learning vision, in particular. AI technology can be applied to countless areas, but only skilled engineers can handle the highly advanced technology for now. We’ve developed a software product designed to help more people, including non-experts, use the difficult technology easily and conveniently so the technology will be used more widely and create greater value.
Deep learning might still be rather unfamiliar to many people. It involves a machine learning method that trains computers to perform human tasks, such as voice recognition, object detection and image classification. Deep learning tech has multiple domains. Neurocle deals with deep learning vision for image analysis, among others.
Deep learning vision technology can be used in various areas. For example, when luggage passes through X-ray scanners at airport security checkpoints, screeners visually inspect every piece of luggage to see if it contains any dangerous substance or items. AI tech can manage this process automatically. Medical doctors use X-rays, CT scans or MRIs to see the internal structures of the body. AI tech can help find any abnormal details that the doctors may miss. In manufacturing, an AI model can automatically identify defective products by simply using the image of the exterior of the goods.
In this era of the COVID-19 pandemic, “contactless” or “non-face-to-face” has been the latest buzzword across the industries, with deep learning technology for image analysis becoming increasingly important.
When a car accident occurs, AI finds out a vehicle type and the extent of the damage just by pictures and automatically calculates estimated repair and compensation costs. Deep learning vision technology can be used in unmanned stores and it can also be applied to the establishment of building entrance security systems. If a person who does not wear a facial mask enters a building, the security system restricts his or her entry after analyzing the face and comparing it to a face wearing a mask. So, automatic contactless entry control is possible.
Amid the rising demand for contact-free services, market demand for this technology has also grown. But it was difficult to actually put the technology into operation. To use deep learning technology, it is necessary to invite engineers with professional knowledge. In reality, however, it was hard to deploy such engineers in every industrial site. Neurocle wondered how to resolve this problem.
Looking back on the mid-1990s, when the Internet started to be widely used, building websites involved the latest technology at the time.
The technology was entirely in the realm of trained engineers who studied computer engineering. But these days, anyone can easily create websites without engineering expertise. I think it’s the process of making technology for general use. At first, only professionals deal with a specific technology. As it’s incorporated into many areas, however, it evolves in a way to be used by anyone. I believe AI technology will go through the same process.
Many companies develop AI software for engineers, because only engineers can use it at present. But our target users are different. We make solutions for non-professionals. In this sense, our software is innovative and clearly different from its alternatives. I’d say we’re taking the lead in allowing many more people to have easier access to deep learning technology.
Neurocle developed its deep learning vision software called “Neuro-T” which trains models for image recognition and “Neuro-R” for a real-time analysis. These two solutions enable even non-experts to build deep learning models. At present, “Neuro-T” and “Neuro-R” are applied to the process of detecting flaws at an auto parts factory and the process of examining imported cosmetics products. They are also used for the analysis of endoscopic images at hospitals and images of microscopic cells at stem cell companies. The use of deep learning vision technology is now expanded to agriculture and education as well.
Image analysis is used in the farming area. When sorting out apples, for example, the apples are classified into different categories, ranging from the highest to the lowest quality. And the prices are set in accordance with the different grades. Generally, people oversee this process. But AI can practice quality grading easily, as long as it is trained.
AI is also a hot topic in the education area. We provide our software to some high schools and colleges as educational material. Students may use the software for practical demonstrations.
For instance, they try to make a self-driving car themselves at a school lab or create a deep learning model that plays the rock-paper-scissors game with an AI model. Like this, our software is consumed in the education market.
In recognition of its technology, Neurocle signed a partnership agreement with a global image data company and made inroads into some 10 countries. This month, it won the Korea Association for ICT Promotion President’s Award at the 2021 Korea ImpaCT-ech Awards to demonstrate the growth potential of its software, “Neuro-T” and “Neuro-R.”
But the company never remains complacent about the achievements. It is determined to move ahead more aggressively with its goal of popularizing AI technology.
In Korea, we have quickly expanded our business in five industries—medical biology, manufacturing, logistics, security and education—since last year. Of course, we’re greatly interested in overseas markets and we’ve already exported our solutions to some Asian countries. When the pandemic is over, we’ll explore bigger markets like Europe and America. I’m pretty sure that our products will be well received in the global market because there is hardly any software for non-professionals. I’ll tell you what. The performance of a deep learning model created by a non-specialist using our software is 2-3 percent higher than that of a similar model made by an engineer working at a large company. In other words, even a non-expert can create a better model than that of a deep learning engineer. This proves our high technology. This year and next, we plan to advance into overseas markets in the five industries I mentioned earlier.
Neurocle develops easy-to-use software that allows even those without professional expertise to use AI technology.
In doing so, the company facilitates the wider use of excellent technology that enriches our lives and provides practical help to self-employed people and small businesses affected by the burden of labor costs. Upholding the value of common prosperity, it continues to move forward.