Kaist

PICASSO Technique Drives Biological Molecules into..

The new imaging approach brings current imaging colors from four to more than 15 for mapping overlapping proteins   Pablo Picasso’s surreal cubist artistic style shifted common features into unrecognizable scenes, but a new imaging approach bearing his namesake may elucidate the most complicated subject: the brain. Employing artificial intelligence to clarify spectral color blending of tiny molecules used to stain specific proteins and other items of research interest, the PICASSO technique, allows researchers to use more than 15 colors to image and parse our overlapping proteins. The PICASSO developers, based in Korea, published their approach on May 5 in Nature Communications.       < 45-color multiplexed imaging of the mouse hippocampus via PICASSO in three staining and imaging rounds (Nature Communications/KAIST) >   Fluorophores — the staining molecules — emit specific colors when excited by a light, but if more than four fluorophores are used, their emitted colors overlap and blend. Researchers previously developed techniques to correct this spectral overlap by precisely defining the matrix of mixed and unmixed images. This measurement depends on reference spectra, found by identifying clear images of only one fluorophore-stained specimen or of multiple, identically prepared specimens that only contain a single fluorophore each.   “Such reference spectra measurement could be complicated to perform in highly heterogeneous specimens, such as the brain, due to the highly varied emission spectra of fluorophores depending on the subregions from which the spectra were measured,” said co-corresponding author Young-Gyu Yoon, professor in the School of Electrical Engineering at KAIST. He explained that the subregions would each need their own spectra reference measurements, making for an inefficient, time-consuming process. “To address this problem, we developed an approach that does not require reference spectra measurements.”   The approach is the “Process of ultra-multiplexed Imaging of biomolecules viA the unmixing of the Signals of Spectrally Overlapping fluorophores,” also known as PICASSO. Ultra-multiplexed imaging refers to visualizing the numerous individual components of a unit. Like a cinema multiplex in which each theater plays a different movie, each protein in a cell has a different role. By staining with fluorophores, researchers can begin to understand those roles.   “We devised a strategy based on information theory; unmixing is performed by iteratively minimizing the mutual information between mixed images,” said co-corresponding author Jae-Byum Chang, professor in the Department of Materials Science and Engineering, KAIST. “This allows us to get away with the assumption that the spatial distribution of different proteins is mutually exclusive and enables accurate information unmixing.”   To demonstrate PICASSO’s capabilities, the researchers applied the technique to imaging a mouse brain. With a single round of staining, they performed 15-color multiplexed imaging of a mouse brain. Although small, mouse brains are still complex, multifaceted organs that can take significant resources to map. According to the researchers, PICASSO can improve the capabilities of other imaging techniques and allow for the use of even more fluorophore colors.   Using one such imaging technique in combination with PICASSO, the team achieved 45-color multiplexed imaging of the mouse brain in only three staining and imaging cycles, according to Yoon.   “PICASSO is a versatile tool for the multiplexed biomolecule imaging of cultured cells, tissue slices and clinical specimens,” Chang said. “We anticipate that PICASSO will be useful for a broad range of applications for which biomolecules’ spatial information is important. One such application the tool would be useful for is revealing the cellular heterogeneities of tumor microenvironments, especially the heterogeneous populations of immune cells, which are closely related to cancer prognoses and the efficacy of cancer therapies.”   The Samsung Research Funding & Incubation Center for Future Technology supported this work. Spectral imaging was performed at the Korea Basic Science Institute Western Seoul Center.     -Publication Junyoung Seo, Yeonbo Sim, Jeewon Kim, Hyunwoo Kim, In Cho, Hoyeon Nam, Yong-Gyu Yoon, Jae-Byum Chang, “PICASSO allows ultra-multiplexed fluorescence imaging of spatially overlapping proteins without reference spectra measurements,” May 5, Nature Communications (doi.org/10.1038/s41467-022-30168-z)     -Profile Professor Jae-Byum Chang Department of Materials Science and Engineering College of Engineering KAIST   Professor Young-Gyu Yoon School of Electrical Engineering College of Engineering KAIST

Professor Juho Kim’s Team Wins Best Paper Award at..

The research team led by Professor Juho Kim from the KAIST School of Computing won a Best Paper Award and an Honorable Mention Award at the Association for Computing Machinery Conference on Human Factors in Computing Systems (ACM CHI) held between April 30 and May 6.   ACM CHI is the world’s most recognized conference in the field of human computer interactions (HCI), and is ranked number one out of all HCI-related journals and conferences based on Google Scholar’s h-5 index. Best paper awards are given to works that rank in the top one percent, and honorable mention awards are given to the top five percent of the papers accepted by the conference.   Professor Juho Kim presented a total of seven papers at ACM CHI 2022, and tied for the largest number of papers. A total of 19 papers were affiliated with KAIST, putting it fifth out of all participating institutes and thereby proving KAIST’s competence in research.       < From left: MS graduate Jeongyeon Kim from the School of Computing, MS candidate Yubin Choi from the School of Electrical Engineering, and Dr. Meng Xia, and Professor Kim >   One of Professor Kim’s research teams composed of Jeongyeon Kim (first author, MS graduate) from the School of Computing, MS candidate Yubin Choi from the School of Electrical Engineering, and Dr. Meng Xia (post-doctoral associate in the School of Computing, currently a post-doctoral associate at Carnegie Mellon University) received a best paper award for their paper, “Mobile-Friendly Content Design for MOOCs: Challenges, Requirements, and Design Opportunities”.   The study analyzed the difficulties experienced by learners watching video-based educational content in a mobile environment and suggests guidelines for solutions. The research team analyzed 134 survey responses and 21 interviews, and revealed that texts that are too small or overcrowded are what mainly brings down the legibility of video contents. Additionally, lighting, noise, and surrounding environments that change frequently are also important factors that may disturb a learning experience.   Based on these findings, the team analyzed the aptness of 41,722 frames from 101 video lectures for mobile environments, and confirmed that they generally show low levels of adequacy. For instance, in the case of text sizes, only 24.5％ of the frames were shown to be adequate for learning in mobile environments. To overcome this issue, the research team suggested a guideline that may improve the legibility of video contents and help overcome the difficulties arising from mobile learning environments.   The importance of and dependency on video-based learning continue to rise, especially in the wake of the pandemic, and it is meaningful that this research suggested a means to analyze and tackle the difficulties of users that learn from the small screens of mobile devices. Furthermore, the paper also suggested technology that can solve problems related to video-based learning through human-AI collaborations, enhancing existing video lectures and improving learning experiences. This technology can be applied to various video-based platforms and content creation.       < From left: Ph.D. candidate Tae Soo Kim, MS candidate DaEun Choi, and Ph.D. candidate Yoonseo Choi, and Professor Kim >   Meanwhile, a research team composed of Ph.D. candidate Tae Soo Kim (first author), MS candidate DaEun Choi, and Ph.D. candidate Yoonseo Choi from the School of Computing received an honorable mention award for their paper, “Stylette: styling the Web with Natural Language”.   The research team developed a novel interface technology that allows nonexperts who are unfamiliar with technical jargon to edit website features through speech. People often find it difficult to use or find the information they need from various websites due to accessibility issues, device-related constraints, inconvenient design, style preferences, etc. However, it is not easy for laymen to edit website features without expertise in programming or design, and most end up just putting up with the inconveniences. But what if the system could read the intentions of its users from their everyday language like “emphasize this part a little more”, or “I want a more modern design”, and edit the features automatically?   Based on this question, Professor Kim’s research team developed ‘Stylette’, a system in which AI analyses its users’ speech expressed in their natural language and automatically recommends a new style that best fits their intentions. The research team created a new system by putting together language AI, visual AI, and user interface technologies. On the linguistic side, a large-scale language model AI converts the intentions of the users expressed through their everyday language into adequate style elements. On the visual side, computer vision AI compares 1.7 million existing web design features and recommends a style adequate for the current website. In an experiment where 40 nonexperts were asked to edit a website design, the subjects that used this system showed double the success rate in a time span that was 35％ shorter compared to the control group.   It is meaningful that this research proposed a practical case in which AI technology constructs intuitive interactions with users. The developed technology can be applied to existing design applications and web browsers in a plug-in format, and can be utilized to improve websites or for advertisements by collecting the natural intention data of users on a large scale.

Now You Can See Floral Scents！​

Optical interferometry visualizes how often lilies emit volatile organic compounds     < Image: Direct measurement of floral scents from a live lily using a laser interferometry method. The signal of the refractive index difference was monitored and the VOCs’ emission frequency results were calculated from Fast Fourier Transform. It showed the unsteady VOC emissions. >   Have you ever thought about when flowers emit their scents?   KAIST mechanical engineers and biological scientists directly visualized how often a lily releases a floral scent using a laser interferometry method. These measurement results can provide new insights for understanding and further exploring the biosynthesis and emission mechanisms of floral volatiles.   Why is it important to know this? It is well known that the fragrance of flowers affects their interactions with pollinators, microorganisms, and florivores. For instance, many flowering plants can tune their scent emission rates when pollinators are active for pollination. Petunias and the wild tobacco Nicotiana attenuata emit floral scents at night to attract night-active pollinators. Thus, visualizing scent emissions can help us understand the ecological evolution of plant-pollinator interactions.   Many groups have been trying to develop methods for scent analysis. Mass spectrometry has been one widely used method for investigating the fragrance of flowers. Although mass spectrometry reveals the quality and quantity of floral scents, it is impossible to directly measure the releasing frequency. A laser-based gas detection system and a smartphone-based detection system using chemo-responsive dyes have also been used to measure volatile organic compounds (VOCs) in real-time, but it is still hard to measure the time-dependent emission rate of floral scents.   However, the KAIST research team co-led by Professor Hyoungsoo Kim from the Department of Mechanical Engineering and Professor Sang-Gyu Kim from the Department of Biological Sciences measured a refractive index difference between the vapor of the VOCs of lilies and the air to measure the emission frequency. The floral scent vapor was detected and the refractive index of air was 1.0 while that of the major floral scent of a linalool lily was 1.46.   Professor Hyoungsoo Kim said, “We expect this technology to be further applicable to various industrial sectors such as developing it to detect hazardous substances in a space.” The research team also plans to identify the DNA mechanism that controls floral scent secretion.   The current work entitled “Real-time visualization of scent accumulation reveals the frequency of floral scent emissions” was published in ‘Frontiers in Plant Science’ on April 18, 2022. (https://doi.org/10.3389/fpls.2022.835305).   This research was supported by the Basic Science Research Program through the National Research Foundation of Korea (NRF-2021R1A2C2007835), the Rural Development Administration (PJ016403), and the KAIST-funded Global Singularity Research PREP-Program.   -Publication: H. Kim, G. Lee, J. Song, and S.-G. Kim, "Real-time visualization of scent accumulation reveals the frequency of floral scent emissions," Frontiers in Plant Science 18, 835305 (2022) (https://doi.org/10.3389/fpls.2022.835305)   -Profile: Professor Hyoungsoo Kim http://fil.kaist.ac.kr ＠MadeInH on Twitter Department of Mechanical Engineering KAIST   Professor Sang-Gyu Kim https://sites.google.com/view/kimlab/home Department of Biological Sciences KAIST

Professor Sang Su Lee’s Team Wins Seven iF Design ..

    Professor Sang Su Lee from the Department of Industrial Design and his team’s five apps made in collaboration with NH Investment and Securities won iF Design Awards in the fields of UI, UX, service design, product design, and communication. These apps are now offered as NH Investment and Securities mobile applications.   The iF Design Awards recognize top quality creativity in product design, communication, packaging, service design and concepts, and architecture and interior design, as well as user experience (UX) and interface for digital media (UI).   In the field of UI, ‘Gretell’ is a mobile stock investment app service designed by Lee and his team to support investors struggling to learn about investing by archiving personalized information. Gretell provides investment information including news and reports. Users learn, evaluate, and leave comments. This shows both quantitative and qualitative indications, leading to rational decision-making. Other user’s comments are shared to reduce confirmation bias. Through this process, Gretell helps users who are impulsive or easily swayed by others’ opinions to grow as independent investors.   ‘Bright’ is another app created by Lee’s team. It helps people exercise their rights as shareholders. As the need to exercise shareholders’ rights increases, many people are frustrated that investors with a small number of shares don’t have a lot of power. Bright provides a space for shareholders to share their opinions and brings people together so that individuals can be more proactive as shareholders. The Integrated Power of Attorney System (IPAS) expands the chances for shareholders to exercise their rights and allows users to submit proposals that can be communicated during the general meeting. Bright fosters influential shareholders, responsible companies, and a healthy society.   For communications, ‘Rewind’ is a stock information services app that visualizes past stock charts through sentiment analysis. Existing services focus on numbers, while Rewind takes a qualitative approach. Rewind analyzes public sentiment toward each event by collecting opinions on social media and then visualizes them chronologically along with the stock chart. Rewind allows users to review stock market movements and record their thoughts. Users can gain their own insights into current events in the stock market and make wiser investment decisions. The intuitive color gradient design provides a pleasant and simplified information experience.   In the area of interfaces for digital media and service design, ‘Groo’ is a green bond investing service app that helps users participate in green investment though investing in green bonds that support green projects for environmental improvement. Not restricted to trading bonds, Groo joins users in the holistic experience of green investing, from taking an interest in environmental issues to confirming the impact of the investment.   Next, ‘Modu’ is a story-based empathy expression training game for children with intellectual disabilities. Modu was developed to support emotion recognition and empathy behavior training in children with mild intellectual disabilities (MID) and borderline intellectual functioning (BIF).   Finally, the diving VR device for neutral buoyancy training, ‘Blow-yancy’, also made winner’s list. The device mimics scuba diving training without having to go into the water, therefore beginner divers are able getting feeling of diving while remaining perfectly safe and not harming any corals. It is expected that the device will be able to help protect at-risk underwater ecosystems.

A New Strategy for Active Metasurface Design Provi..

The new strategy displays an unprecedented upper limit of dynamic phase modulation with no significant variations in optical amplitude   An international team of researchers led by Professor Min Seok Jang of KAIST and Professor Victor W. Brar of the University of Wisconsin-Madison has demonstrated a widely applicable methodology enabling a full 360° active phase modulation for metasurfaces while maintaining significant levels of uniform light amplitude. This strategy can be fundamentally applied to any spectral region with any structures and resonances that fit the bill.   Metasurfaces are optical components with specialized functionalities indispensable for real-life applications ranging from LIDAR and spectroscopy to futuristic technologies such as invisibility cloaks and holograms. They are known for their compact and micro/nano-sized nature, which enables them to be integrated into electronic computerized systems with sizes that are ever decreasing as predicted by Moore’s law.       < Figure1.The metasurface designed by the team that demonstrates complete 2π tunable phase modulation utilizing the avoided crossing of two resonances. >       < Figure2. a: Complex reflection coefficient trajectories with different mobility values for the graphene sheet case. Full 2π phase modulation does not occur without the avoided crossing with graphene plasmons, despite the increasing mobilities and therefore the decreasing linewidths. b: Complex reflection coefficient trajectories with different mobility values for the graphene ribbon case.) >   In order to allow for such innovations, metasurfaces must be capable of manipulating the impinging light, doing so by manipulating either the light’s amplitude or phase (or both) and emitting it back out. However, dynamically modulating the phase with the full circle range has been a notoriously difficult task, with very few works managing to do so by sacrificing a substantial amount of amplitude control.   Challenged by these limitations, the team proposed a general methodology that enables metasurfaces to implement a dynamic phase modulation with the complete 360° phase range, all the while uniformly maintaining significant levels of amplitude.   The underlying reason for the difficulty achieving such a feat is that there is a fundamental trade-off regarding dynamically controlling the optical phase of light. Metasurfaces generally perform such a function through optical resonances, an excitation of electrons inside the metasurface structure that harmonically oscillate together with the incident light. In order to be able to modulate through the entire range of 0-360°, the optical resonance frequency (the center of the spectrum) must be tuned by a large amount while the linewidth (the width of the spectrum) is kept to a minimum. However, to electrically tune the optical resonance frequency of the metasurface on demand, there needs to be a controllable influx and outflux of electrons into the metasurface and this inevitably leads to a larger linewidth of the aforementioned optical resonance.   The problem is further compounded by the fact that the phase and the amplitude of optical resonances are closely correlated in a complex, non-linear fashion, making it very difficult to hold substantial control over the amplitude while changing the phase.   The team’s work circumvented both problems by using two optical resonances, each with specifically designated properties. One resonance provides the decoupling between the phase and amplitude so that the phase is able to be tuned while significant and uniform levels of amplitude are maintained, as well as providing a narrow linewidth.   The other resonance provides the capability of being sufficiently tuned to a large degree so that the complete full circle range of phase modulation is achievable. The quintessence of the work is then to combine the different properties of the two resonances through a phenomenon called avoided crossing, so that the interactions between the two resonances lead to an amalgamation of the desired traits that achieves and even surpasses the full 360° phase modulation with uniform amplitude.   Professor Jang said, “Our research proposes a new methodology in dynamic phase modulation that breaks through the conventional limits and trade-offs, while being broadly applicable in diverse types of metasurfaces. We hope that this idea helps researchers implement and realize many key applications of metasurfaces, such as LIDAR and holograms, so that the nanophotonics industry keeps growing and provides a brighter technological future.”   The research paper authored by Ju Young Kim and Juho Park, et al., and titled "Full 2π Tunable Phase Modulation Using Avoided Crossing of Resonances" was published in Nature Communications on April 19. The research was funded by the Samsung Research Funding & Incubation Center of Samsung Electronics.     -Publication: Ju Young Kim, Juho Park, Gregory R. Holdman, Jacob T. Heiden, Shinho Kim, Victor W. Brar, and Min Seok Jang, “Full 2π Tunable Phase Modulation Using Avoided Crossing of Resonances” Nature Communications on April 19 (2022). doi.org/10.1038/s41467-022-29721-7   -Profile Professor Min Seok Jang School of Electrical Engineering KAIST

LightPC Presents a Resilient System Using Only Non..

Lightweight Persistence Centric System (LightPC) ensures both data and execution persistence for energy-efficient full system persistence         A KAIST research team has developed hardware and software technology that ensures both data and execution persistence. The Lightweight Persistence Centric System (LightPC) makes the systems resilient against power failures by utilizing only non-volatile memory as the main memory.   “We mounted non-volatile memory on a system board prototype and created an operating system to verify the effectiveness of LightPC,” said Professor Myoungsoo Jung. The team confirmed that LightPC validated its execution while powering up and down in the middle of execution, showing up to eight times more memory, 4.3 times faster application execution, and 73％ lower power consumption compared to traditional systems.   Professor Jung said that LightPC can be utilized in a variety of fields such as data centers and high-performance computing to provide large-capacity memory, high performance, low power consumption, and service reliability.   In general, power failures on legacy systems can lead to the loss of data stored in the DRAM-based main memory. Unlike volatile memory such as DRAM, non-volatile memory can retain its data without power. Although non-volatile memory has the characteristics of lower power consumption and larger capacity than DRAM, non-volatile memory is typically used for the task of secondary storage due to its lower write performance. For this reason, nonvolatile memory is often used with DRAM. However, modern systems employing non-volatile memory-based main memory experience unexpected performance degradation due to the complicated memory microarchitecture.   To enable both data and execution persistent in legacy systems, it is necessary to transfer the data from the volatile memory to the non-volatile memory. Checkpointing is one possible solution. It periodically transfers the data in preparation for a sudden power failure. While this technology is essential for ensuring high mobility and reliability for users, checkpointing also has fatal drawbacks. It takes additional time and power to move data and requires a data recovery process as well as restarting the system.   In order to address these issues, the research team developed a processor and memory controller to raise the performance of non-volatile memory-only memory. LightPC matches the performance of DRAM by minimizing the internal volatile memory components from non-volatile memory, exposing the non-volatile memory (PRAM) media to the host, and increasing parallelism to service on-the-fly requests as soon as possible.   The team also presented operating system technology that quickly makes execution states of running processes persistent without the need for a checkpointing process. The operating system prevents all modifications to execution states and data by keeping all program executions idle before transferring data in order to support consistency within a period much shorter than the standard power hold-up time of about 16 minutes. For consistency, when the power is recovered, the computer almost immediately revives itself and re-executes all the offline processes immediately without the need for a boot process.   The researchers will present their work (LightPC: Hardware and Software Co-Design for Energy-Efficient Full System Persistence) at the International Symposium on Computer Architecture (ISCA) 2022 in New York in June. More information is available at the CAMELab website (http://camelab.org).       < Figure 1. Customized system prototype (left: System prototype, Right: Implementation view) >       < Figure 2. Overview of the proposed LightPC >       < Figure 3. Energy and performance of the LightPC/LegacyPC (Right: Cycle norm. to legacy PC) >     -Profile: Professor Myoungsoo Jung Computer Architecture and Memory Systems Laboratory (CAMEL) http://camelab.org School of Electrical Engineering KAIST

Energy-Efficient AI Hardware Technology Via a Brai..

Researchers demonstrate neuromodulation-inspired stashing system for the energy-efficient learning of a spiking neural network using a self-rectifying memristor array       < Image: A schematic illustrating the localized brain activity (a-c) and the configuration of the hardware and software hybrid neural network (d-e) using a self-rectifying memristor array (f-g). >     Researchers have proposed a novel system inspired by the neuromodulation of the brain, referred to as a ‘stashing system,’ that requires less energy consumption. The research group led by Professor Kyung Min Kim from the Department of Materials Science and Engineering has developed a technology that can efficiently handle mathematical operations for artificial intelligence by imitating the continuous changes in the topology of the neural network according to the situation. The human brain changes its neural topology in real time, learning to store or recall memories as needed. The research group presented a new artificial intelligence learning method that directly implements these neural coordination circuit configurations.   Research on artificial intelligence is becoming very active, and the development of artificial intelligence-based electronic devices and product releases are accelerating, especially in the Fourth Industrial Revolution age. To implement artificial intelligence in electronic devices, customized hardware development should also be supported. However most electronic devices for artificial intelligence require high power consumption and highly integrated memory arrays for large-scale tasks. It has been challenging to solve these power consumption and integration limitations, and efforts have been made to find out how the human brain solves problems.   To prove the efficiency of the developed technology, the research group created artificial neural network hardware equipped with a self-rectifying synaptic array and algorithm called a ‘stashing system’ that was developed to conduct artificial intelligence learning. As a result, it was able to reduce energy by 37％ within the stashing system without any accuracy degradation. This result proves that emulating the neuromodulation in humans is possible.   Professor Kim said, "In this study, we implemented the learning method of the human brain with only a simple circuit composition and through this we were able to reduce the energy needed by nearly 40 percent.”   This neuromodulation-inspired stashing system that mimics the brain’s neural activity is compatible with existing electronic devices and commercialized semiconductor hardware. It is expected to be used in the design of next-generation semiconductor chips for artificial intelligence.   This study was published in Advanced Functional Materials in March 2022 and supported by KAIST, the National Research Foundation of Korea, the National NanoFab Center, and SK Hynix.     -Publication: Woon Hyung Cheong, Jae Bum Jeon†, Jae Hyun In, Geunyoung Kim, Hanchan Song, Janho An, Juseong Park, Young Seok Kim, Cheol Seong Hwang, and Kyung Min Kim (2022) “Demonstration of Neuromodulation-inspired Stashing System for Energy-efficient Learning of Spiking Neural Network using a Self-Rectifying Memristor Array,” Advanced Functional Materials March 31, 2022 (DOI: 10.1002/adfm.202200337)   -Profile: Professor Kyung Min Kim http://semi.kaist.ac.kr https://scholar.google.com/citations?user=BGw8yDYAAAAJ&hl=ko Department of Materials Science and Engineering KAIST

Machine Learning-Based Algorithm to Speed up DNA S..

The algorithm presents the first full-fledged, short-read alignment software that leverages learned indices for solving the exact match search problem for efficient seeding       < Image:Scientists from KAIST develops new machine-learning-based approach to speed up DNA sequencing. >   The human genome consists of a complete set of DNA, which is about 6.4 billion letters long. Because of its size, reading the whole genome sequence at once is challenging. So scientists use DNA sequencers to produce hundreds of millions of DNA sequence fragments, or short reads, up to 300 letters long. Then the DNA sequencer assembles all the short reads like a giant jigsaw puzzle to reconstruct the entire genome sequence. Even with very fast computers, this job can take hours to complete.   A research team at KAIST has achieved up to 3.45x faster speeds by developing the first short-read alignment software that uses a recent advance in machine-learning called a learned index.   The research team reported their findings on March 7, 2022 in the journal Bioinformatics. The software has been released as open source and can be found on github (https://github.com/kaist-ina/BWA-MEME).   Next-generation sequencing (NGS) is a state-of-the-art DNA sequencing method. Projects are underway with the goal of producing genome sequencing at population scale. Modern NGS hardware is capable of generating billions of short reads in a single run. Then the short reads have to be aligned with the reference DNA sequence. With large-scale DNA sequencing operations running hundreds of next-generation sequences, the need for an efficient short read alignment tool has become even more critical. Accelerating the DNA sequence alignment would be a step toward achieving the goal of population-scale sequencing. However, existing algorithms are limited in their performance because of their frequent memory accesses.   BWA-MEM2 is a popular short-read alignment software package currently used to sequence the DNA. However, it has its limitations. The state-of-the-art alignment has two phases – seeding and extending. During the seeding phase, searches find exact matches of short reads in the reference DNA sequence. During the extending phase, the short reads from the seeding phase are extended. In the current process, bottlenecks occur in the seeding phase. Finding the exact matches slows the process.   The researchers set out to solve the problem of accelerating the DNA sequence alignment. To speed the process, they applied machine learning techniques to create an algorithmic improvement. Their algorithm, BWA-MEME (BWA-MEM emulated) leverages learned indices to solve the exact match search problem. The original software compared one character at a time for an exact match search. The team’s new algorithm achieves up to 3.45x faster speeds in seeding throughput over BWA-MEM2 by reducing the number of instructions by 4.60x and memory accesses by 8.77x. “Through this study, it has been shown that full genome big data analysis can be performed faster and less costly than conventional methods by applying machine learning technology,” said Professor Dongsu Han from the School of Electrical Engineering at KAIST.   The researchers’ ultimate goal was to develop efficient software that scientists from academia and industry could use on a daily basis for analyzing big data in genomics. “With the recent advances in artificial intelligence and machine learning, we see so many opportunities for designing better software for genomic data analysis. The potential is there for accelerating existing analysis as well as enabling new types of analysis, and our goal is to develop such software,” added Han.   Whole genome sequencing has traditionally been used for discovering genomic mutations and identifying the root causes of diseases, which leads to the discovery and development of new drugs and cures. There could be many potential applications. Whole genome sequencing is used not only for research, but also for clinical purposes. “The science and technology for analyzing genomic data is making rapid progress to make it more accessible for scientists and patients. This will enhance our understanding about diseases and develop a better cure for patients of various diseases.”   The research was funded by the National Research Foundation of the Korean government’s Ministry of Science and ICT.     -Publication Youngmok Jung, Dongsu Han, “BWA-MEME:BWA-MEM emulated with a machine learning approach,” Bioinformatics, Volume 38, Issue 9, May 2022 (https://doi.org/10.1093/bioinformatics/btac137)   -Profile Professor Dongsu Han School of Electrical Engineering KAIST

Professor June-Koo Rhee’s Team Wins the QHack Open..

  < From left: Ju-young Ryu, Jeung-rak Lee, and Eyuel Elala in Professor June-Koo Rhee >     The research team consisting of three master students Ju-Young Ryu, Jeung-rak Lee, and Eyel Elala in Professor June-Koo Rhee’s group from the KAIST IRTC of Quantum Computing for AI has won the first place at the QHack 2022 Open Hackathon Science Challenge.   The QHack 2022 Open Hackathon is one of the world’s prestigious quantum software hackathon events held by US Xanadu, in which 250 people from 100 countries participate. Major sponsors such as IBM Quantum, AWS, CERN QTI, and Google Quantum AI proposed challenging problems, and a winning team is selected judged on team projects in each of the 13 challenges.   The KAIST team supervised by Professor Rhee received the First Place prize on the Science Challenge which was organized by the CERN QTI of the European Communities. The team will be awarded an opportunity to tour CERN’s research lab in Europe for one week along with an online internship.   The students on the team presented a method for “Leaning Based Error Mitigation for VQE,” in which they implemented an LBEM protocol to lower the error in quantum computing, and leveraged the protocol in the VQU algorithm which is used to calculate the ground state energy of a given molecule.   Their research successfully demonstrated the ability to effectively mitigate the error in IBM Quantum hardware and the virtual error model. In conjunction, Professor June-Koo (Kevin) Rhee founded a quantum computing venture start-up, Qunova Computing(https://qunovacomputing.com), with technology tranfer from the KAIST ITRC of Quantum Computing for AI. Qunova Computing is one of the frontier of the quantum software industry in Korea.

Decoding Brain Signals to Control a Robotic Arm​

Advanced brain-machine interface system successfully interprets arm movement directions from neural signals in the brain       < Figure:Experimental paradigm. Subjects were instructed to perform reach-and-grasp movements to designate the locations of the target in three-dimensional space. (a) Subjects A and B were provided the visual cue as a real tennis ball at one of four pseudo-randomized locations. (b) Subjects A and B were provided the visual cue as a virtual reality clip showing a sequence of five stages of a reach-and-grasp movement. >   Researchers have developed a mind-reading system for decoding neural signals from the brain during arm movement. The method, described in the journal Applied Soft Computing, can be used by a person to control a robotic arm through a brain-machine interface (BMI).   A BMI is a device that translates nerve signals into commands to control a machine, such as a computer or a robotic limb. There are two main techniques for monitoring neural signals in BMIs: electroencephalography (EEG) and electrocorticography (ECoG).   The EEG exhibits signals from electrodes on the surface of the scalp and is widely employed because it is non-invasive, relatively cheap, safe and easy to use. However, the EEG has low spatial resolution and detects irrelevant neural signals, which makes it difficult to interpret the intentions of individuals from the EEG.   On the other hand, the ECoG is an invasive method that involves placing electrodes directly on the surface of the cerebral cortex below the scalp. Compared with the EEG, the ECoG can monitor neural signals with much higher spatial resolution and less background noise. However, this technique has several drawbacks.   “The ECoG is primarily used to find potential sources of epileptic seizures, meaning the electrodes are placed in different locations for different patients and may not be in the optimal regions of the brain for detecting sensory and movement signals,” explained Professor Jaeseung Jeong, a brain scientist at KAIST. “This inconsistency makes it difficult to decode brain signals to predict movements.”   To overcome these problems, Professor Jeong’s team developed a new method for decoding ECoG neural signals during arm movement. The system is based on a machine-learning system for analysing and predicting neural signals called an ‘echo-state network’ and a mathematical probability model called the Gaussian distribution.   In the study, the researchers recorded ECoG signals from four individuals with epilepsy while they were performing a reach-and-grasp task. Because the ECoG electrodes were placed according to the potential sources of each patient’s epileptic seizures, only 22％ to 44％ of the electrodes were located in the regions of the brain responsible for controlling movement.   During the movement task, the participants were given visual cues, either by placing a real tennis ball in front of them, or via a virtual reality headset showing a clip of a human arm reaching forward in first-person view. They were asked to reach forward, grasp an object, then return their hand and release the object, while wearing motion sensors on their wrists and fingers. In a second task, they were instructed to imagine reaching forward without moving their arms.   The researchers monitored the signals from the ECoG electrodes during real and imaginary arm movements, and tested whether the new system could predict the direction of this movement from the neural signals. They found that the novel decoder successfully classified arm movements in 24 directions in three-dimensional space, both in the real and virtual tasks, and that the results were at least five times more accurate than chance. They also used a computer simulation to show that the novel ECoG decoder could control the movements of a robotic arm.   Overall, the results suggest that the new machine learning-based BCI system successfully used ECoG signals to interpret the direction of the intended movements. The next steps will be to improve the accuracy and efficiency of the decoder. In the future, it could be used in a real-time BMI device to help people with movement or sensory impairments.   This research was supported by the KAIST Global Singularity Research Program of 2021, Brain Research Program of the National Research Foundation of Korea funded by the Ministry of Science, ICT, and Future Planning, and the Basic Science Research Program through the National Research Foundation of Korea funded by the Ministry of Education.   -Publication Hoon-Hee Kim, Jaeseung Jeong, “An electrocorticographic decoder for arm movement for brain-machine interface using an echo state network and Gaussian readout,” Applied Soft Computing online December 31, 2021 (doi.org/10.1016/j.asoc.2021.108393)   -Profile Professor Jaeseung Jeong Department of Bio and Brain Engineering College of Engineering KAIST

AI Light-Field Camera Reads 3D Facial Expressions​

Machine-learned, light-field camera reads facial expressions from high-contrast illumination invariant 3D facial images       < Image: Facial expression reading based on MLP classification from 3D depth maps and 2D images obtained by NIR-LFC >   A joint research team led by Professors Ki-Hun Jeong and Doheon Lee from the KAIST Department of Bio and Brain Engineering reported the development of a technique for facial expression detection by merging near-infrared light-field camera techniques with artificial intelligence (AI) technology.   Unlike a conventional camera, the light-field camera contains micro-lens arrays in front of the image sensor, which makes the camera small enough to fit into a smart phone, while allowing it to acquire the spatial and directional information of the light with a single shot. The technique has received attention as it can reconstruct images in a variety of ways including multi-views, refocusing, and 3D image acquisition, giving rise to many potential applications.   However, the optical crosstalk between shadows caused by external light sources in the environment and the micro-lens has limited existing light-field cameras from being able to provide accurate image contrast and 3D reconstruction.   The joint research team applied a vertical-cavity surface-emitting laser (VCSEL) in the near-IR range to stabilize the accuracy of 3D image reconstruction that previously depended on environmental light. When an external light source is shone on a face at 0-, 30-, and 60-degree angles, the light field camera reduces 54％ of image reconstruction errors. Additionally, by inserting a light-absorbing layer for visible and near-IR wavelengths between the micro-lens arrays, the team could minimize optical crosstalk while increasing the image contrast by 2.1 times.   Through this technique, the team could overcome the limitations of existing light-field cameras and was able to develop their NIR-based light-field camera (NIR-LFC), optimized for the 3D image reconstruction of facial expressions. Using the NIR-LFC, the team acquired high-quality 3D reconstruction images of facial expressions expressing various emotions regardless of the lighting conditions of the surrounding environment.   The facial expressions in the acquired 3D images were distinguished through machine learning with an average of 85％ accuracy – a statistically significant figure compared to when 2D images were used. Furthermore, by calculating the interdependency of distance information that varies with facial expression in 3D images, the team could identify the information a light-field camera utilizes to distinguish human expressions.   Professor Ki-Hun Jeong said, “The sub-miniature light-field camera developed by the research team has the potential to become the new platform to quantitatively analyze the facial expressions and emotions of humans.” To highlight the significance of this research, he added, “It could be applied in various fields including mobile healthcare, field diagnosis, social cognition, and human-machine interactions.”   This research was published in Advanced Intelligent Systems online on December 16, under the title, “Machine-Learned Light-field Camera that Reads Facial Expression from High-Contrast and Illumination Invariant 3D Facial Images.” This research was funded by the Ministry of Science and ICT and the Ministry of Trade, Industry and Energy.   -Publication “Machine-learned light-field camera that reads fascial expression from high-contrast and illumination invariant 3D facial images,” Sang-In Bae, Sangyeon Lee, Jae-Myeong Kwon, Hyun-Kyung Kim. Kyung-Won Jang, Doheon Lee, Ki-Hun Jeong, Advanced Intelligent Systems, December 16, 2021 (doi.org/10.1002/aisy.202100182)   -Profile Professor Ki-Hun Jeong Biophotonic Laboratory Department of Bio and Brain Engineering KAIST   Professor Doheon Lee Department of Bio and Brain Engineering KAIST

Face Detection in Untrained Deep Neural Networks​

A KAIST team shows that primitive visual selectivity of faces can arise spontaneously in completely untrained deep neural networks         Researchers have found that higher visual cognitive functions can arise spontaneously in untrained neural networks. A KAIST research team led by Professor Se-Bum Paik from the Department of Bio and Brain Engineering has shown that visual selectivity of facial images can arise even in completely untrained deep neural networks.   This new finding has provided revelatory insights into mechanisms underlying the development of cognitive functions in both biological and artificial neural networks, also making a significant impact on our understanding of the origin of early brain functions before sensory experiences.   The study published in Nature Communications on December 16 demonstrates that neuronal activities selective to facial images are observed in randomly initialized deep neural networks in the complete absence of learning, and that they show the characteristics of those observed in biological brains.   The ability to identify and recognize faces is a crucial function for social behavior, and this ability is thought to originate from neuronal tuning at the single or multi-neuronal level. Neurons that selectively respond to faces are observed in young animals of various species, and this raises intense debate whether face-selective neurons can arise innately in the brain or if they require visual experience.   Using a model neural network that captures properties of the ventral stream of the visual cortex, the research team found that face-selectivity can emerge spontaneously from random feedforward wirings in untrained deep neural networks. The team showed that the character of this innate face-selectivity is comparable to that observed with face-selective neurons in the brain, and that this spontaneous neuronal tuning for faces enables the network to perform face detection tasks.   These results imply a possible scenario in which the random feedforward connections that develop in early, untrained networks may be sufficient for initializing primitive visual cognitive functions.   Professor Paik said, “Our findings suggest that innate cognitive functions can emerge spontaneously from the statistical complexity embedded in the hierarchical feedforward projection circuitry, even in the complete absence of learning”.   He continued, “Our results provide a broad conceptual advance as well as advanced insight into the mechanisms underlying the development of innate functions in both biological and artificial neural networks, which may unravel the mystery of the generation and evolution of intelligence.” This work was supported by the National Research Foundation of Korea (NRF) and by the KAIST singularity research project.   -Publication Seungdae Baek, Min Song, Jaeson Jang, Gwangsu Kim, and Se-Bum Baik, “Face detection in untrained deep neural network,” Nature Communications 12, 7328 on Dec.16, 2021 (https://doi.org/10.1038/s41467-021-27606-9)   -Profile Professor Se-Bum Paik Visual System and Neural Network Laboratory Program of Brain and Cognitive Engineering Department of Bio and Brain Engineering College of Engineering KAIST