KAIST Reveals Mathematical Principle behind AI’s ‘..
(from left: Professor Jong Chul Ye, PhD candidates Yoseob Han and Eunju Cha) A KAIST research team identified the geometrical structure of artificial intelligence (AI) and discovered the mathematical principles of highly performing artificial neural networks, which can be applicable in fields such as medical imaging. Deep neural networks are an exemplary method of implementing deep learning, which is at the core of the AI technology, and have shown explosive growth in recent years. This technique has been used in various fields, such as image and speech recognition as well as image processing. Despite its excellent performance and usefulness, the exact working principles of deep neural networks has not been well understood, and they often suffer from unexpected results or errors. Hence, there is an increasing social and technical demand for interpretable deep neural network models. To address these issues, Professor Jong Chul Ye from the Department of Bio & Brain Engineering and his team attempted to find the geometric structure in a higher dimensional space where the structure of the deep neural network can be easily understood. They proposed a general deep learning framework, called deep convolutional framelets, to understand the mathematical principle of a deep neural network in terms of the mathematical tools in Harmonic analysis. As a result, it was found that deep neural networks’ structure appears during the process of decomposition of high dimensionally lifted signal via Hankel matrix, which is a high-dimensional structure formerly studied intensively in the field of signal processing. In the process of decomposing the lifted signal, two bases categorized as local and non-local basis emerge. The researchers found that non-local and local basis functions play a role in pooling and filtering operation in convolutional neural network, respectively. Previously, when implementing AI, deep neural networks were usually constructed through empirical trial and errors. The significance of the research lies in the fact that it provides a mathematical understanding on the neural network structure in high dimensional space, which guides users to design an optimized neural network. They demonstrated improved performance of the deep convolutional framelets’ neural networks in the applications of image denoising, image pixel in painting, and medical image restoration. Professor Ye said, “Unlike conventional neural networks designed through trial-and-error, our theory shows that neural network structure can be optimized to each desired application and are easily predictable in their effects by exploiting the high dimensional geometry. This technology can be applied to a variety of fields requiring interpretation of the architecture, such as medical imaging.” This research, led by PhD candidates Yoseob Han and Eunju Cha, was published in the April 26th issue of the SIAM Journal on Imaging Sciences. Figure 1. The design of deep neural network using mathematical principles Figure 2. The results of image noise cancelling Figure 3. The artificial neural network restoration results in the case where 80％ of the pixels are lost
Professor Emeritus Jung Ki Park Won the IBA Techno..
(Professor Emeritus Jung Ki Park) Professor Emeritus Jung Ki Park from the Department of Chemical and Biomolecular Engineering received the IBA Technology Award from the International Battery Association (IBA). IBA 2018 was held from March 11 to 16 on Jeju Island, which was the first time it was hosted in Korea. The conference was an excellent opportunity to let the world know the level of the Korean rechargeable battery industry and its technology. Professor Park delivered his keynote speech titled Advances in Lithium Batteries in Korea at the conference and received the IBA Technology Award as the first Korean recipient. ?Professor Park is a world-renowned scholar who was a groundbreaker in the rechargeable battery industry. He was recognized by the IBA Award Committee for his contributions carrying out research and development, fostering competent people, and enhancing the lithium rechargeable battery industry in Korea over the last 30 years. Professor Park said, “It is my great honor to receive this award, which is the best international award in the field of rechargeable batteries. I would like to share this with my colleagues and students. As competition in the rechargeable industry intensifies, systemic cooperation among industries, academia, and government is needed for the continued development of the battery industry in Korea.
Seong-Tae Kim Wins Robert-Wagner All-Conference Be..
(Ph.D. candidate Seong-Tae Kim) Ph.D. candidate Seong-Tae Kim from the School of Electrical Engineering won the Robert Wagner All-Conference Best Student Paper Award during the 2018 International Society for Optics and Photonics (SPIE) Medical Imaging Conference, which was held in Houston last month. Kim, supervised by Professor Yong Man Ro, received the award for his paper in the category of computer-aided diagnosis. His paper, titled “ICADx: Interpretable Computer-Aided Diagnosis of Breast Masses”, was selected as the best paper out of 900 submissions. The conference selects the best paper in nine different categories. His research provides new insights on diagnostic technology to detect breast cancer powered by deep learning.
Recognizing Seven Different Face Emotions on a Mob..
(Professor Hoi-Jun Yoo) A KAIST research team succeeded in achieving face emotion recognition on a mobile platform by developing an AI semiconductor IC that processes two neural networks on a single chip. Professor Hoi-Jun Yoo and his team (Primary researcher: Jinmook Lee Ph. D. student) from the School of Electrical Engineering developed a unified deep neural network processing unit (UNPU). Deep learning is a technology for machine learning based on artificial neural networks, which allows a computer to learn by itself, just like a human. The developed chip adjusts the weight precision (from 1 bit to 16 bit) of a neural network inside of the semiconductor in order to optimize energy efficiency and accuracy. With a single chip, it can process a convolutional neural network (CNN) and recurrent neural network (RNN) simultaneously. CNN is used for categorizing and recognizing images while RNN is for action recognition and speech recognition, such as time-series information. Moreover, it enables an adjustment in energy efficiency and accuracy dynamically while recognizing objects. To realize mobile AI technology, it needs to process high-speed operations with low energy, otherwise the battery can run out quickly due to processing massive amounts of information at once. According to the team, this chip has better operation performance compared to world-class level mobile AI chips such as Google TPU. The energy efficiency of the new chip is 4 times higher than the TPU. In order to demonstrate its high performance, the team installed UNPU in a smartphone to facilitate automatic face emotion recognition on the smartphone. This system displays a user’s emotions in real time. The research results for this system were presented at the 2018 International Solid-State Circuits Conference (ISSCC) in San Francisco on February 13. Professor Yoo said, “We have developed a semiconductor that accelerates with low power requirements in order to realize AI on mobile platforms. We are hoping that this technology will be applied in various areas, such as object recognition, emotion recognition, action recognition, and automatic translation. Within one year, we will commercialize this technology.”
KAIST Professors Selected as Y-KAST Members
Professor YongKeun Park, Professor Bumjoon Kim, Professor Keon Jae Lee, and Professor Young Seok Ju were selected as the newest members of the Young Korean Academy of Science and Technology (Y-KAST).The Korean Academy of Science and Technology, an academic institution of professional experts, selected 26 promising scientists under the age of 43 to join Y-KAST. and four KAIST professors were included in the list.The newest members were conferred on February 26.Research FieldNameNatural SciencesYongKeun Park (Dept. of Physics)Engineering Bumjoon Kim (Dept. of Chemical and Biomolecular Engineering)Agricultural & Fishery SciencesKeon Jae Lee (Dept. of Materials Science and Engineering)Medical SciencesYoung Seok Ju (Graduate School of Medical Science and Engineering)
KAIST, First to Win the Cube Satellite Competition
?Professor Hyochoong Bang from the Department of Aerospace Engineering and his team received the Minister of Science and ICT Award at the 1st Cube Satellite Competition. The team actually participated in the competition in 2012, but it took several years for the awarding ceremony since it took years for the satellites to be designed, produced, and launched. The KAIST team successfully developed a cube satellite, named ‘Little Intelligent Nanosatellite of KAIST (LINK)’ and completed its launch in April 2017. LINK (size: 20cmx10cmx10cm, weight: 2kg) mounted mass spectrometry and Langmuir probe for Earth observation. The Langmuir probe was developed by Professor Kyoung Wook Min from the Department of Physics, KAIST. Yeerang Lim, a PhD student from the Department of Aerospace Engineering said, “I still remember the feeling that I had on the day when LINK launched into orbit and sent back signals. I hope that space exploration is not something far away but attainable for us in near future.”
Professor Il-Doo Kim, the 20th Recipient of the So..
Professor Il-Doo Kim from the Department of Materials Science and Engineering at KAIST received the 20th Song-gok Science and Technology Award from Korea Institute of Science and Technology (KSIT).The Song-gok Science and Technology Award was established to praise the accomplishments of the first president, Hyung-seop Choi, whose penname is Song-gok. The award selects a recipient in the field of materials and technology every other year.Professor Kim, in recognition of his outstanding research and contributions to materials science in Korea, received the award during the 52nd anniversary ceremony of KIST on February 9.Professor Kim focuses on developing nanofiber gas sensors for diagnosing disease in advance by analyzing exhaled biomarkers with electrospinning technology.He has published more than 211 papers and has recorded more than 9,650 citations and 50 h-index.Professor Kim has registered 107 patents and applied 38 patents in Korea while registering 29 patents and applying 16 patents overseas. Also, he transferred four technologies in 2017.Professor Kim is recognized as one of the researchers who is leading nanofiber technology. On January 17, he made a keynote speech at the 5th International Conference on Electrospinning, which was his fourth keynote speech at that conference.Moreover, he received the Technology Innovation Award at the College of Engineering, KAIST on December 19, 2017.Professor Kim said, “It is my great honor to receive the Song-gok Science and Technology Award. I would like to bring distinction to KAIST by taking the lead in the commercializing a nanofiber-based highly sensitive nanosensors, diversifying and commercializing technology using nanofiber.”
First Woman to Receive Grand Prize of Samsung Huma..
Yeunhee Huh, PhD candidate (Professor Gyu-Hyeong Cho) from the School of Electrical Engineering received the grand prize of the 24th Humantech Paper Award. She is the first female recipient of this prize since its establishment in 1994. The Humantech Paper Award is hosted by Samsung Electronics and sponsored by the Ministry of Science and ICT with JoongAng Daily Newspaper. Her paper is titled, ‘A Hybrid Structure Dual-Path Step-Down Converter with 96.2％ Peak Efficiency using 250mΩ Large-DCR Inductor’. Electronic devices require numerous chips and have a power converter to supply energy adequately. She proposed a new structure to enhance energy efficiency by combining inductors and capacitors. Enhancing energy efficiency can reduce energy loss, which prolongs battery hours and solves overheating of devices; for instance, energy loss leads to the overheating issue affecting phone chargers. This technology can be applied to various electronic devices, such as cell phones, laptops, and drones. Huh said, “Power has to go up in order to meet customers’ needs; however the overheating problem emerges during this process. This problem affects surrounding circuits and causes other issues, such as malfunctions of electronic devices. This technology may vary according to the conditions, but it can enhance energy efficiency up to 4%.”During the ceremony, about eight hundred million KRW worth cash prizes was conferred to 119 papers. KAIST (44 papers) and Gyeonggi Science High School (6 papers) received special awards given to the schools.
Professor Jungwon Kim Wins the 8th Haerim Optics a..
(Professor Jungwon Kim) Professor Jungwon Kim from the Department of Mechanical Engineering received the 8th Haerim Optics and Photonics Award from the Optical Society of Korea (OSK). He was recognized for his dedication to pioneering the field of microwave photonics by developing ultra-low noise fiber photonics lasers. The Haerim Optics and Photonics Award is given to an outstanding researcher who has made academic contributions in the field of optics and photonics for the last five years. The name of the award (Haerim) comes from the pen-name of the renowned scholar, Professor Un-Chul Paek, because it is maintained using funds he contributed to the OSK. The OSK will confer the award on February 8 during the 29th OSK Annual Meeting and Winter Conference of 2018.
Three Professors Named KAST Fellows
(Professor Dan Keun Sung at the center) (Professor Y.H. Cho at the center) (Professor K.H. Cho at the center) The Korean Academy of Science and Technology (KAST) inducted three KAIST professors as fellows at the New Year’s ceremony held at KAST on January 12. They were among the 24 newly elected fellows of the most distinguished academy in Korea. The new fellows are Professor Dan Keun Sung of the School of Electrical Engineering, Professor Kwang-Hyun Cho of the Department of Bio and Brain Engineering, and Professor Yong-Hoon Cho of the Department of Physics. Professor Sung was recognized for his lifetime academic achievements in fields related with network protocols and energy ICT. He also played a crucial role in launching the Korean satellites KITSAT-1,2,3 and the establishment of the Satellite Technology Research Center at KAIST. Professor Y.H.Cho has been a pioneer in the field of low-dimensional semiconductor-powered quantum photonics that enables quantum optical research in solid state. He has been recognized as a renowned scholar in this field internationally. Professor K.H.Cho has conducted original research that combines IT and BT in systems biology and has applied novel technologies of electronic modeling and computer simulation analysis for investigating complex life sciences. Professor Cho, who is in his 40s, is the youngest fellow among the newly inducted fellows.
Professor Jung Awarded the Pople Medal by the APAT..
(Professor Yousung Jung) Professor Yousung Jung of the Graduate School of EEWS won the Pople Medal from the Asia-Pacific Association of Theoretical & Computational Chemists (APATCC). The Pople Medal has been awarded annually since 2007 to recognize young scholars in the fields of theoretical/computational chemistry in honor of Sir John Anthony Pople, who passed away in 2004. Dr. Pople was a British theoretical chemist and a Nobel laureate in 1998 for his development of computational methods in quantum chemistry. The Pople Medal is awarded to scientists at or under the age of 45 in the Asia-Pacific region who have distinguished themselves through pioneering and important contributions. Professor Jung was honored for his outstanding contributions to developing efficient electronic structure methods and their applications to energy materials discovery. He has published more than 120 papers in prestigious academic journals. He also has an h-index of 44, and has been cited more than 8,000 times.
Distinguished Professor Sang Yup Lee Named NAI Fel..
(Distinguished Professor Sang Yup Lee) Distinguished Professor Sang Yup Lee of the Department of Chemical and Biomolecular Engineering was named to the National Academy of Inventors in the US. He is the first Korean scholar ever elected as a NAI fellow. The NAI is a non-profit member organization with over 4,000 individual inventors and fellows spanning more than 250 institutions worldwide. It is comprised of universities as well as governmental and non-profit research institutes. The academy was founded in 2010 to recognize and encourage inventors with patents from the US Patent and Trademark Office. So far, 575 fellows from 229 institutions have been elected. The academy said Professor Lee has been recognized for fellowship induction as he has demonstrated a highly prolific spirit of innovation in creating or facilitating outstanding inventions that have made a tangible impact on quality of life, economic development, and the welfare of society. Distinguished Professor Lee, a pioneering researcher and scholar in the field of systems metabolic engineering, was ranked in the top 1％ of highly cited researchers (HCR) this year. Over the past 11 years, he published more than 130,000 articles in prestigious journals around the world. He has been cited more than 34,000 times since he started working at KAIST in 1994. He is also the first Korean ever elected to both the National Academy of Sciences (NAS) and the National Academy of Engineering (NAE) in the US, becoming the one of 13 foreign scholars in the world holding two prestigious institutions’ fellowships. Dr. Lee is currently the dean of KAIST Institutes, the world-leading institute for multi and interdisciplinary research. He is also serving as co-chair of the Global Council on Biotechnology and is a member of the Global Future Council on the Fourth Industrial Revolution at the World Economic Forum.