EE Prof. Junmo Kim’s Team, Develop AI That Imagines and Understands How Images Change Like Humans

김준모 교수 유재명 박사과정 증명사진
<Professor Junmo Kim, PhD candidate Jaemyung Yu>

A research team led by Professor Junmo Kim from the Department of Electrical Engineering at KAIST has developed an innovative AI technology that can envision and understand how images change, similar to how humans imagine transformations like rotation or recoloring. This breakthrough goes beyond simply analyzing images, enabling the AI to comprehend and express the processes involved in transforming visual data. The technology holds promise for diverse applications, including medical imaging, autonomous driving, and robotics, where precision and adaptability are essential.

 

AI That Imagine Changes Like Humans (Understands How Images Change, Like Humans)

 

The newly developed technology, Self-supervised Transformation Learning (STL), focuses on enabling AI to learn how images transform. STL operates without relying on human-provided labels; instead, it learns transformations by comparing original images with their transformed versions. It independently recognizes changes such as, “This has been rotated,” or, “The color has changed.” This process parallels the way humans observe, imagine, and interpret variations in visual data.

 

Illustration of the roles of the three representation learning approaches that constitute STL
Illustration of the roles of the three representation learning approaches that constitute STL: (a) distinguishing images regardless of transformations, (b) aligning transformation representations for the same transformation applied to different images, and (c) ensuring that relationships between representations of transformed variants of the same image reflect the actual transformation. STL integrates all these roles for comprehensive learning

 

Overcoming the Limitations of Conventional Methods

 

Traditional AI systems often struggle with subtle transformations, focusing primarily on capturing large, overarching features while ignoring finer details. This limitation becomes a significant challenge in scenarios where precise understanding of intricate changes is crucial.

 

STL addresses this gap by learning to encode even the smallest transformations in an image into its feature space—a conceptual map representing the relationships between different data points. Rather than ignoring these changes, STL incorporates them into its feature representations, enabling more accurate and nuanced outcomes.

 

For example, STL excels at recognizing specific alterations, such as random cropping, brightness adjustments, and color modifications, achieving performance improvements of up to 42% over conventional methods. It is particularly adept at handling complex transformations that were previously difficult for AI to manage.

 

Examples of transformations often ignored by existing methods
Examples of transformations often ignored by existing methods. These transformations, though subtle, may carry important differences.

 

Smarter AI for Broader Applications

 

What sets STL apart is its ability to not only understand visual content but also learn and represent transformations themselves. This capability allows STL to detect subtle changes in medical images, such as CT scans, and better interpret diverse conditions in autonomous driving. By incorporating transformations into its understanding, STL can deliver safer and more precise results across various applications.

 

Toward Human-Like Understanding

 

“STL represents a significant leap forward in AI technology, closely mirroring the way humans perceive and interpret changes in images,” said Professor Junmo Kim. “This approach has the potential to drive innovations in fields such as healthcare, robotics, and self-driving cars, where understanding transformations is critical.”

 

The research, conducted by Jaemyung Yu, a PhD candidate at KAIST as the first author, was presented at NeurIPS 2024, one of the world’s leading AI conferences, under the title Self-supervised Transformation Learning for Equivariant Representations. It was supported by the Ministry of Science and ICT through the Institute of Information and Communications Technology Planning and Evaluation (IITP) as part of the SW StarLab program (No. RS-2024-00439020, Development of Sustainable Real-time Multimodal Interactive Generative AI).

EE Professor Chan-Hyun Youn’s Research Team Develops an Unlearning Method Using DNN-Based Weight Prediction Technique

241212 윤찬현 교수님 main
<(from left) Professor Chan-Hyun Youn, Ph.D. candidate Jinhyuk Jang>

 

Professor Chan-Hyun Youn’s research team has developed a novel approach to Machine Unlearning, a critical aspect of AI safety, by introducing the past weight prediction model InvWNN. This model aims to selectively remove the influence of problematic data from AI models trained on such data. Existing methods often require access to the entire training dataset or face challenges with performance degradation. To address these issues, the team proposed a new method that leverages weight history to predict past weights and iteratively applies this process to progressively eliminate the influence of the data.

 

241212 main2

<Figure 1: Machine Unlearning Process of the Proposed InvWNN and Unlearning Trajectory>

 

This method gradually removes the problematic data’s influence through iterative fine-tuning on the problematic data and the past weight prediction. Notably, this approach operates effectively without access to the remaining data and can be applied to various datasets and architectures. Compared to existing methods, the proposed technique excels at accurately removing unnecessary knowledge from training data while minimizing side effects. Furthermore, it has been validated that this method can be directly applied to a variety of tasks without requiring additional procedures.

 

The research team demonstrated the high performance of their method across various benchmarks, showcasing its potential to significantly expand the practical applicability of machine unlearning technology. These findings will be presented at the 39th Annual AAAI Conference on Artificial Intelligence (AAAI 2025), one of the premier international conferences in the field of artificial intelligence, to be held in the United States in February next year, with the title “Learning to Rewind via Iterative Prediction of Past Weights for Unlearning.”

 

Electrical Engineering Professor Junil Choi Receives the 6th Next-Generation Scientist Award

최준일 교수 시상식 사진
<Professor Junil Choi>

 

Professor Junil Choi of our School of Electrical Engineering has been awarded the 6th Next-Generation Scientist Award (IT field), co-hosted by the Korean Academy of Science and Technology (KAST) and the Korean Association of University Presidents, and sponsored by the S-OIL Science and Culture Foundation.

 

The Korean Academy of Science and Technology (https://kast.or.kr) has been running the S-OIL Awards program since 2011 with support from the S-OIL Science and Culture Foundation. This initiative identifies and recognizes young scientists in the fields of basic science and engineering to foster them as key contributors to national science and technology in the 21st century.

 

The S-OIL Next-Generation Scientist Award is presented to researchers aged 45 or younger. Candidates are evaluated based on their top 10 representative research papers published over the last 10 years, excluding those from their doctoral studies or postdoctoral periods.

 

Professor Junil Choi received the award on December 5, 2024, in recognition of his outstanding research in communication systems leveraging machine learning and his contributions to 6G wireless communication systems.

 

The School of Electrical Engineering at KAIST has demonstrated remarkable consistency, with Professors Hyun-Joo Lee (2022), Jae-Woong Jeong (2023), and now Junil Choi (2024) receiving the Next-Generation Scientist Award in the IT field for three consecutive years.

EE Prof. Hyunjoo Jenny Lee’s Lab’s Eunyoung Jang and Kiup Kim were awarded the Best Paper Award in graduate students’ poster presentations at the ‘2024 Micro and Nano Systems Fall Conference’

장은영 김기업 사진
<(from left) Ph.D. candidates Eunyoung Jang and Kiup Kim>

Ph.D. candidates Eunyoung Jang and Kiup Kim from Professor Hyunjoo Jenny Lee’s laboratory in our department were awarded the Best Paper Award in graduate students’ poster presentations at the ‘2024 Micro and Nano Systems Fall Conference’. This event was held on November 21 at the Asti Hotel in Busan.
 
The ‘Micro and Nano Systems Fall Conference’ has been an annual event since 2020, aimed at fostering academic and personal exchanges.
 
The title of the winning paper is ‘3D Organoid Multi-functional Monitoring Platforms for Real-time and Non-invasive Analysis’. The paper, contributed to by both Eunyoung Jang and Kiup Kim, was recognized for its excellence within the Medical MEMS category, one of the four academic divisions at the conference.
 

EE Prof. Hyunjoo Lee, Receives Young Researcher Award from the Korean Society of Therapeutic Ultrasound

이현주 교수 인물 사진

 

Professor of Electrical Engineering Hyunjoo Lee has been honored with the Young Researcher Award at the 2024 Korean Society for Therapeutic Ultrasound (KSTU) conference.

 

This prestigious award recognizes outstanding research achievements in the field of therapeutic ultrasound and is granted to one researcher under the age of 43. The Korean Society for Therapeutic Ultrasound is an academic organization dedicated to the advancement of therapeutic ultrasound technology through academic research and innovation. Its initiatives include promoting research and clinical applications of diagnostic and therapeutic ultrasound technologies while fostering interdisciplinary collaboration and exchange among related fields.

 

Therapeutic ultrasound, in particular, is a groundbreaking technology that uses ultrasound energy to stimulate tissues or treat pathological conditions. It has drawn significant attention for its innovative applications in various medical domains, including cancer treatment, neural stimulation, and thrombolysis. By strengthening the academic foundation of therapeutic ultrasound, the society aims to facilitate collaboration between researchers and clinicians, contributing to advancements in medical technology and improving patient quality of life.

 

Professor Lee has been conducting research on neural interfaces and brain-computer interfaces, focusing on developing non-invasive brain stimulation technologies using ultrasound for bidirectional neural interfaces and the treatment of neurological disorders. This award acknowledges Professor Lee’s contributions to the field of therapeutic ultrasound and highlights her potential as a next-generation leader in the domain.

EE Prof. SangHyeon Kim’s research team, Secures Core Technology for Ultra-High-Resolution Image Sensors​

images 000085 photo1.jpg 7
<Photo. (From left) Professor SangHyeon Kim, Professor Dae-Myung Geum of Inha University (formerly a KAIST postdoctoral researcher) and Dr. Jinha Lim (currently a postdoctoral researcher at Yale University)

 

A joint research team from Korea and the United States has developed next-generation, high-resolution image sensor technology with higher power efficiency and a smaller size compared to existing sensors. Notably, they have secured foundational technology for ultra-high-resolution shortwave infrared (SWIR) image sensors, an area currently dominated by Sony, paving the way for future market entry.

 

Research team led by Professor SangHyeon Kim from the School of Electrical Engineering, in collaboration with Inha University and Yale University in the U.S., has developed an ultra-thin broadband photodiode (PD), marking a significant breakthrough in high-performance image sensor technology.

 

This research drastically improves the trade-off between the absorption layer thickness and quantum efficiency found in conventional photodiode technology. Specifically, it achieved high quantum efficiency of over 70% even in an absorption layer thinner than one micrometer (μm), reducing the thickness of the absorption layer by approximately 70% compared to existing technologies.

 

A thinner absorption layer simplifies pixel processing, allowing for higher resolution and smoother carrier diffusion, which is advantageous for light carrier acquisition while also reducing the cost. However, a fundamental issue with thinner absorption layers is the reduced absorption of long-wavelength light. 

 

images 000086 image01 900
< Figure 1. Schematic diagram of the InGaAs photodiode image sensor integrated on the Guided-Mode Resonance (GMR) structure proposed in this study (left), a photograph of the fabricated wafer, and a scanning electron microscope (SEM) image of the periodic patterns (right) >

 

The research team introduced a guided-mode resonance (GMR) structure* that enables high-efficiency light absorption across a wide spectral range from 400 nanometers (nm) to 1,700 nanometers (nm). This wavelength range includes not only visible light but also light the SWIR region, making it valuable for various industrial applications. *Guided-Mode Resonance (GMR) Structure: A concept used in electromagnetics, a phenomenon in which a specific (light) wave resonates (forming a strong electric/magnetic field) at a specific wavelength. Since energy is maximized under these conditions, it has been used to increase antenna or radar efficiency.

 

The improved performance in the SWIR region is expected to play a significant role in developing next-generation image sensors with increasingly high resolutions. The GMR structure, in particular, holds potential for further enhancing resolution and other performance metrics through hybrid integration and monolithic 3D integration with complementary metal-oxide-semiconductor (CMOS)-based readout integrated circuits (ROIC).

 

images 000086 image02 900
< Figure 2. Benchmark for state-of-the-art InGaAs-based SWIR pixels with simulated EQE lines as a function of TAL variation. Performance is maintained while reducing the absorption layer thickness from 2.1 micrometers or more to 1 micrometer or less while reducing it by 50% to 70% >

 

The research team has significantly enhanced international competitiveness in low-power devices and ultra-high-resolution imaging technology, opening up possibilities for applications in digital cameras, security systems, medical and industrial image sensors, as well as future ultra-high-resolution sensors for autonomous driving, aerospace, and satellite observation.

 

Professor Sang Hyun Kim, the lead researcher, commented, “This research demonstrates that significantly higher performance than existing technologies can be achieved even with ultra-thin absorption layers.” 

 

images 000086 image03 900
< Figure 3. Top optical microscope image and cross-sectional scanning electron microscope image of the InGaAs photodiode image sensor fabricated on the GMR structure (left). Improved quantum efficiency performance of the ultra-thin image sensor (red) fabricated with the technology proposed in this study (right) >

 

The results of this research were published on 15th of November, in the prestigious international journal Light: Science & Applications (JCR 2.9%, IF=20.6), with Professor Dae-Myung Geum of Inha University (formerly a KAIST postdoctoral researcher) and Dr. Jinha Lim (currently a postdoctoral researcher at Yale University) as co-first authors. (Paper title: “Highly-efficient (>70%) and Wide-spectral (400 nm -1700 nm) sub-micron-thick InGaAs photodiodes for future high-resolution image sensors”)

 

This study was supported by the National Research Foundation of Korea.

EE Professor Ian Oakley’s Lab’s Jiwan Kim and Hoheon Jeong Win Best People’s Choice Award at the ACM UIST Student Innovation Contest

EE Professor Ian Oakley’s Lab’s Jiwan Kim and Hoheon Jeong

Win Best People’s Choice Award at the ACM UIST Student Innovation Contest

< (From left) Jiwan Kim (PhD candidate), Hoheon Jung (undergraduate) >

 

PhD candidate Jiwan Kim, and undergraduate Hoheon Jung from Professor Ian Oakley’s lab in the Department of Electrical Engineering, won the Best People’s Choice Award at the Student Innovation Contest held as part of the ‘ACM UIST (ACM Symposium on User Interface Software and Technology)’ in Pittsburgh, USA, from October 13 to 16. The Best People’s Choice Award is given to the project that garners the most enthusiasm and support from attendees during the conference.

 

672c1ce581ba4
<Awarded certificate and trophy>

The ‘ACM UIST’ is a leading international conference in the field of human-computer interaction. Each year, the Student Innovation Contest invites teams to present innovative ideas using cutting-edge hardware just before its release. This year’s theme involved creating and demonstrating interactive devices for the future using the Gen-M Kit from Seeed Studio. After a competitive preliminary round, eight teams from prestigious institutions, including Carnegie Mellon University, the University of Toronto, and the University of Hong Kong, reached the finals alongside our university’s team.

 

Jiwan Kim and Hoheon Jung cited the famous quote by novelist Arthur C. Clarke, known for works like 2001: A Space Odyssey: “Any sufficiently advanced technology is indistinguishable from magic.” Inspired by this idea, they developed a wearable device that provides an experience similar to superpowers.

The glove they developed uses surface acoustic waves, radar, and ultrasound to create features such as eavesdropping to hear sounds through walls, enhanced senses to detect nearby movements with closed eyes, and telekinesis to levitate small objects in the air.

Jiwan Kim remarked, “Some might view this as simply implementing technology for amusement, but I believe that fun is also an essential direction for scientific and technological advancement. We focused on interpreting various sensing technologies in the most entertaining way possible and demonstrating them accordingly.”

EE Prof. Jung-Yong Lee research team developed a high-efficiency and high-stability organic-inorganic hybrid solar cell production technology that maximizes near-infrared light capture

 

images 000084 photo1.jpg 3

 <Photo. (From left) Professor Jung-Yong Lee, Ph.D. candidate Min-Ho Lee, and Master’s candidate Min Seok Kim of the School of Electrical Engineering>

 

Existing perovskite solar cells, which have the problem of not being able to utilize approximately 52% of total solar energy, have been developed by a Korean research team as an innovative technology that maximizes near-infrared light capture performance while greatly improving power conversion efficiency. This greatly increases the possibility of commercializing next-generation solar cells and is expected to contribute to important technological advancements in the global solar cell market.

 

The research team of Professor Jung-Yong Lee of the KAIST EE and Professor Woojae Kim of the Department of Chemistry at Yonsei University developed a high-efficiency and high-stability organic-inorganic hybrid solar cell production technology that maximizes near-infrared light capture beyond the existing visible light range.

 

The research team suggested and advanced a hybrid next-generation device structure with organic photo-semiconductors that complements perovskite materials limited to visible light absorption and expands the absorption range to near-infrared.

 

In addition, they revealed the electronic structure problem that mainly occurs in the structure and announced a high-performance solar cell device that dramatically solved this problem by introducing a dipole layer*. *Dipole layer: A thin material layer that controls the energy level within the device to facilitate charge transport and forms an interface potential difference to improve device performance.

 

Existing lead-based perovskite solar cells have a problem in that their absorption spectrum is limited to the visible light region with a wavelength of 850 nanometers (nm) or less, which prevents them from utilizing approximately 52% of the total solar energy.

 

To solve this problem, the research team designed a hybrid device that combined an organic bulk heterojunction (BHJ) with perovskite and implemented a solar cell that can absorb up to the near-infrared region.

 

In particular, by introducing a sub-nanometer dipole interface layer, they succeeded in alleviating the energy barrier between the perovskite and the organic bulk heterojunction (BHJ), suppressing charge accumulation, maximizing the contribution to the near-infrared, and improving the current density (JSC) to 4.9 mA/cm².

 

The key achievement of this study is that the power conversion efficiency (PCE) of the hybrid device has been significantly increased from 20.4% to 24.0%. In particular, this study achieved a high internal quantum efficiency (IQE) compared to previous studies, reaching 78% in the near-infrared region.

 

images 000084 Eng image01 1 900

< Figure. The illustration of the mechanism of improving the electronic structure and charge transfer capability through Perovskite/organic hybrid device structure and dipole interfacial layers (DILs). The proposed dipole interfacial layer forms a strong interfacial dipole, effectively reducing the energy barrier between the perovskite and organic bulk heterojunction (BHJ), and suppressing hole accumulation. This technology improves near-infrared photon harvesting and charge transfer, and as a result, the power conversion efficiency of the solar cell increases to 24.0%. In addition, it achieves excellent stability by maintaining performance for 1,200 hours even in an extremely humid environment. >

 

In addition, this device showed high stability, showing excellent results of maintaining more than 80% of the initial efficiency in the maximum output tracking for more than 800 hours even under extreme humidity conditions.

 

Professor Jung-Yong Lee said, “Through this study, we have effectively solved the charge accumulation and energy band mismatch problems faced by existing perovskite/organic hybrid solar cells, and we will be able to significantly improve the power conversion efficiency while maximizing the near-infrared light capture performance, which will be a new breakthrough that can solve the mechanical-chemical stability problems of existing perovskites and overcome the optical limitations.”

 

This study, in which KAIST School of Electrical Engineering Ph.D. candidate Min-Ho Lee and Master’s candidate Min Seok Kim participated as co-first authors, was published in the September 30th online edition of the international academic journal Advanced Materials. (Paper title: Suppressing Hole Accumulation Through Sub-Nanometer Dipole Interfaces in Hybrid Perovskite/Organic Solar Cells for Boosting Near-Infrared Photon Harvesting).

 

This study was conducted with the support of the National Research Foundation of Korea.

EE Prof. Yong-Hoon Kim’s team succeeded in accelerating calculations for electronic structure in quantum mechanics for the first time in the world using a convolutional neural network (CNN) model

images 000084 photo1.jpg 2

< (from left):  Prof. Yong-Hoon Kim, Ph.D. candidate  Ryong Gyu Lee>

 

The close relationship between AI and highly complicated scientific computing can be seen in the fact that both the 2024 Nobel Prizes in Physics and Chemistry were awarded to scientists for devising the AI for their respective fields of study. KAIST researchers succeeded in dramatically reducing the computation time for highly sophisticated computer simulations for quantum mechanics by predicting atomic-level chemical bonding information distributed in 3D space using a novel approach to teach AI.

 

Professor Yong-Hoon Kim’s team from the School of Electrical Engineering developed a 3D computer vision artificial neural network-based computation methodology that bypasses the complex algorithms required for atomic-level quantum mechanical calculations traditionally performed using supercomputers to derive the properties of materials.

 

images 000084 Image 1 900 2

< Figure 1. Various methodologies are utilized in the simulation of materials and materials, such as quantum mechanical calculations at the nanometer (nm) level, classical mechanical force fields at the scale of tens to hundreds of nanometers, continuum dynamics calculations at the macroscopic scale, and calculations that mix simulations at different scales. These simulations are already playing a key role in a wide range of basic research and application development fields in combination with informatics techniques. Recently, there have been active efforts to introduce machine learning techniques to radically accelerate simulations, but research on introducing machine learning techniques to quantum mechanical electronic structure calculations, which form the basis of high-scale simulations, is still insufficient. >

 

The density functional theory (DFT) calculations in quantum mechanics using supercomputers have become an essential and standard tool in a wide range of research and development fields, including advanced materials and drug design, as they allow for fast and accurate prediction of quantum properties. *Density functional theory (DFT): A representative theory of ab initio (first principles) calculations that calculate quantum mechanical properties from the atomic level.

 

However, practical DFT calculations require generating 3D electron density and solving quantum mechanical equations through a complex, iterative self-consistent field (SCF)* process that must be repeated tens to hundreds of times. This restricts its application to systems with only a few hundred to a few thousand atoms. *Self-consistent field (SCF): A scientific computing method widely used to solve complex many-body problems that must be described by a number of interconnected simultaneous differential equations.

 

Professor Yong-Hoon Kim’s research team questioned whether recent advancements in AI techniques could be used to bypass the SCF process. As a result, they developed the DeepSCF model, which accelerates calculations by learning chemical bonding information distributed in a 3D space using neural network algorithms from the field of computer vision.

 

images 000084 Image 2 900

< Figure 2. The deepSCF methodology developed in this study provides a way to rapidly accelerate DFT calculations by avoiding the self-consistent field process (orange box) that had to be performed repeatedly in traditional quantum mechanical electronic structure calculations through artificial neural network techniques (green box). The self-consistent field process is a process of predicting the 3D electron density, constructing the corresponding potential, and then solving the quantum mechanical Cohn-Sham equations, repeating tens to hundreds of times. The core idea of the deepSCF methodology is that the residual electron density (δρ), which is the difference between the electron density (ρ) and the sum of the electron densities of the constituent atoms (ρ0), corresponds to chemical bonding information, so the self-consistent field process is replaced with a 3D convolutional neural network model. >

 

The research team focused on the fact that, according to density functional theory, electron density contains all quantum mechanical information of electrons, and that the residual electron density — the difference between the total electron density and the sum of the electron densities of the constituent atoms — contains chemical bonding information. They used this as the target for machine learning.

 

They then adopted a dataset of organic molecules with various chemical bonding characteristics, applying random rotations and deformations to the atomic structures of these molecules to further enhance the model’s accuracy and generalization capabilities. Ultimately, the research team demonstrated the validity and efficiency of the DeepSCF methodology on large, complex systems.

 

images 000084 image3.jpg

< Figure 3. An example of applying the deepSCF methodology to a carbon nanotube-based DNA sequence analysis device model (top left). In addition to classical mechanical interatomic forces (bottom right), the residual electron density (top right) and quantum mechanical electronic structure properties such as the electronic density of states (DOS) (bottom left) containing information on chemical bonding are rapidly predicted with an accuracy corresponding to the standard DFT calculation results that perform the SCF process. >

 

Professor Yong-Hoon Kim, who supervised the research, explained that his team had found a way to map quantum mechanical chemical bonding information in a 3D space onto artificial neural networks. He noted, “Since quantum mechanical electron structure calculations underpin property simulations at all scales, this research establishes a foundational principle for accelerating material calculations using artificial intelligence.”

 

Ryong-Gyu Lee, a PhD candidate in the School of Electrical Engineering, served as the first author of this research, which was published online on October 24 in Npj Computational Materials, a prestigious journal in the field of material computation. (Paper title: “Convolutional network learning of self-consistent electron density via grid-projected atomic fingerprints”)

 

This research was conducted with support from the KAIST Venture Research Program for Graduate and PhD Students and the National Research Foundation of Korea’s Mid-career Researcher Support Program.

EE Ph.D. candidate Edward Jongyoon Choi and Vincent Lukito from Professor Minkyu Je’s Lab Wins Corporate Special Award (Telechips) at the 25th Korea Semiconductor Design Challenge

EE Ph.D. candidate Edward Jongyoon Choi and Vincent Lukito from Professor Minkyu Je’s Lab

Wins Corporate Special Award (Telechips) at the 25th Korea Semiconductor Design Challenge

e1730271496130

 <(From left) Ph.D. candidate Edward Jongyoon Choi and Vincent Lukito>

 

Ph.D. candidate Edward Jongyoon Choi and Vincent Lukito from Professor Minkyu Je’s lab in EE was awarded the prestigious Corporate Special Award (Telechips) at the 25th Korea Semiconductor Design Challenge, held on October 24 at COEX, Seoul.

 

Organized by the Ministry of Trade, Industry and Energy and the Korea Semiconductor Industry Association, the Korea Semiconductor Design Challenge aims to cultivate the design capabilities of undergraduate and graduate students in the semiconductor field and to discover creative ideas that enhance the foundational competitiveness of the semiconductor industry.

 

대전 수상 사진 1

  <Edward Jongyoon Choi and Vincent Lukito at the awards ceremony>

 

The title of their award-winning research design is “Spike Sorting SoC with Delta-based Detection and Analog CIM-based Autoencoder Neural Network Feature Extraction Achieving 94.54% Accuracy,” in which both Ph.D. students participated.

 

The research was evaluated based on creativity, technical complexity, commercial viability, and completeness. Their project demonstrated outstanding merit in terms of innovative topics, high technical difficulty and excellence, potential for commercialization, and the completeness and validation of the work, receiving the Special Corporate Award (Telechips).