-
- [Jan 12, 2026] Professor Hyungjoon Koo SecAI Lab, a paper accepted for FSE '26 NEW
- The paper titled "Fool Me If You Can: On the Robustness of Binary Code Similarity Detection Models against Semantics-preserving Transformations," co-authored by Ji-yong Eom (Ph.D. candidate), Min-seok Kim (M.S. candidate), both from the SecAI Lab (supervised by Hyungjoon Koo, https://secai.skku.edu/), and Michalis Polychronakis from Stony Brook University, has been accepted for publication at the prestigious Foundations of Software Engineering 2026 (FSE '26). The paper will be presented in July 2026. Software reverse engineering is a critical process in the security field, including vulnerability analysis and malware detection, but it requires a high level of expertise. However, relying solely on such methods presents limitations in effectively addressing the rapidly increasing modern threats. To overcome this challenge, recent approaches have actively proposed techniques to assist reverse engineering with artificial intelligence, especially models that extract contextual information from machine code (assembly language). Similar to how natural language can convey meaning through context-preserving transformations, assembly language also has techniques for transforming code while maintaining the same semantics (semantics-preserving code transformations). However, there has been a lack of in-depth analysis on how robust artificial intelligence models are against these types of transformations. This study systematically analyzes the impact of eight transformation techniques on the performance of six representative AI-based binary similarity detection models. It also introduces how models can lead to incorrect judgments, such as false positives and false negatives. For this, a dataset consisting of 9,565 transformed binaries from 620 original binaries was built for experimentation. The results show that the robustness to transformations varies based on the architecture and preprocessing methods of the models, and that even slight transformations can significantly degrade model performance, especially if the attacker designs the transformation precisely. This research emphasizes that, when designing AI models for supporting reverse engineering, model robustness against binary transformations should be considered as crucial as performance metrics. Abstract: Binary code analysis plays an essential role in cybersecurity, facilitating reverse engineering to reveal the inner workings of programs in the absence of source code. Traditional approaches, such as static and dynamic analysis, extract valuable insights from stripped binaries, but often demand substantial expertise and manual effort. Recent advances in deep learning have opened promising opportunities to enhance binary analysis by capturing latent features and disclosing underlying code semantics. Despite the growing number of binary analysis models based on machine learning, their robustness to adversarial code transformations at the binary level remains underexplored to date. In this work, we evaluate the robustness of deep learning models for the task of binary code similarity detection (BCSD) under semantics-preserving transformations. The unique nature of machine instructions presents distinct challenges compared to the typical input perturbations found in other domains. To achieve our goal, we introduce asmFooler, a system that evaluates the resilience of BCSD models using a diverse set of adversarial code transformations that preserve functional semantics. We construct a dataset of 9,565 binary variants from 620 baseline samples by applying eight semantics-preserving transformations across six representative BCSD models. Our major findings highlight several key insights: i) model robustness highly relies on the design of the processing pipeline, including code pre-processing, model architecture, and internal feature selection, which collectively determine how code semantics are captured; ii) the effectiveness of adversarial transformations is bounded by a transformation budget, shaped by model-specific constraints such as input size limits and the expressive capacity of semantically equivalent instructions; iii) well-crafted adversarial transformations can be highly effective, even when introducing minimal perturbations; and iv) such transformations efficiently disrupt the model's decision (e.g., misleading to false positives or false negatives) by focusing on semantically significant instructions. | Professor Hyungjoon Koo | kevin.koo@skku.edu, kevinkoo001.github.io | SecAI Lab | secai.skku.edu/
-
- 작성일 2026-01-12
- 조회수 56
-
- [Dec 30, 2025] SKKU Produces Winners of the 2025 Korea Talent Award (Sangho Kim, School of Convergence)
- SKKU Produces Winners of the 2025 Korea Talent Award (Sangho Kim, School of Convergence) SKKU Has Four Students Selected as Winners of the 2025 Korea Talent Award in the University and Youth Categories (From left to right) Sangho Kim, Seungmin Kim, Jinhyung Lee, Yoonseo Heo A total of four students from our university have been selected as winners of the '2025 Korea Talent Award' in the University and Youth General Category, proving the university’s success in nurturing talents who excel in both academic achievements and social contributions. The '2025 Korea Talent Award,' organized by the Ministry of Education and managed by the Korea Scholarship Foundation, selected a total of 100 recipients, including 40 high school students and 60 university students and adults, after going through regional and central evaluations. This award was established to discover and foster future talents who can create new values based on creativity and passion, and contribute to the development of the community. ▲(From left to right) Awardees Seungmin Kim and Jinhyung Lee The selected students from Sungkyunkwan University are making meaningful achievements based on expertise and public service in their respective fields of study and activities. Sangho Kim, from the School of Convergence, has contributed to solving social issues through research and practical activities bridging academic and public sectors. Seungmin Kim, from the Department of Pharmacy, has continued his public contribution activities by connecting academic achievements based on his major with societal returns. Jinhyung Lee, from the Department of Mechanical Engineering, has accumulated significant research achievements by linking research with practical applications. Yoonseo Heo, from the Department of Sports Science, is showing great potential as a future talent in her field by balancing academic and professional activities. ▲(From left to right) Awardee Yoonseo Heo and Sangho Kim's certificate This award is seen as a demonstration of our university's ongoing success in nurturing well-rounded talents who possess both academic competence and social responsibility, in line with our educational philosophy of 'Self-Cultivation and Governance of Others (修己治人).' Moving forward, our university plans to continue fostering holistic talents who will lead the future of the nation and society by linking education, research, and social contributions. ▲ 2025 Korea Talent Award group photo
-
- 작성일 2026-01-05
- 조회수 132
-
- [Dec 29, 2025] Professor Jaehoon Paul Jeong receives Minister of Science and ICT Award for leading international Inte... NEW
- Recipient of the Minister of Science and ICT Award for Leading International Internet Standardization Professor Jaehoon Paul Jeong of the Department of Computer Science and Engineering, College of Computing and Informatics, received the Minister of Science and ICT Award at the Global ICT Standards Conference (GISC) 2025 (https://gisc.or.kr/) held at EL Tower in Seoul on November 3. Since its launch in 2017, GISC has been the largest international event in Korea focused on standards and patents, bringing together experts to discuss the future of next-generation ICT technologies and standards. GISC 2025, held under the theme “AI for All,” gathered ICT standardization professionals, industry representatives, academics, and research institutions from around the world to share the latest knowledge and insights. The conference focused on standardization issues in key and emerging technologies such as AI, 6G, quantum technologies, and digital transformation, with global companies, international standardization organizations, and policy agencies participating to discuss interoperability, reliability, and the integration of standards and intellectual property, presenting a vision for the future digital society. Each year, GISC recognizes researchers who have made significant contributions to the development and standardization of international technologies by awarding the Minister of Science and ICT Award. Professor Jaehoon Paul Jeong was recognized at GISC 2025 for his contributions in developing cloud-based security service technologies and vehicle networking technologies for autonomous vehicles. These technologies were implemented in multiple international standards approved by the Internet Engineering Task Force (IETF), and further shared through open-source projects to demonstrate proof of concept (POC), earning him the Minister of Science and ICT Award for his contributions to ICT standardization in 2025. Figure 1. Receiving the Minister of Science and ICT Award for ICT Standard Figure 2. Certificate of the Minister of Science and ICT Award for ICT Standard Professor Jaehoon Paul Jeong has been engaged in Internet technology standardization at the Internet Engineering Task Force (IETF) for 23 years, from 2002 to 2025, and continues to actively participate as an ICT international standardization expert at TTA (Telecommunications Technology Association of Korea). As shown in Figure 3, Professor Paul co-authored RFC 8192 on problem statements and use cases for the I2NSF (Interface to Network Security Functions) Working Group (WG) for cloud-based security systems. He also served as editor for six I2NSF WG documents that have been approved as RFCs. These six documents are scheduled to be published as RFCs in the first half of 2026. The SKKU IoT Lab team led by Professor Paul (http://iotlab.skku.edu/) participated in IETF hackathons with the I2NSF Framework Project, winning four awards (IETF-97, IETF-99, IETF-100, IETF-103), contributing to global recognition of Korea's Internet technologies in the I2NSF field. Professor Paul also edited the problem statements and use case documents for the IPWAVE (IP Wireless Access in Vehicular Environments) WG for IPv6-based vehicular networking, which were published as RFC 9365. To implement proof of concept (POC) for I2NSF WG and IPWAVE WG standard documents, he leads open-source projects on GitHub. Over the past 23 years (2002–2025), he has served as lead author of two RFCs, RFC 4339 and RFC 5006 (later updated as RFC 6106 and RFC 8106). He actively participates in standardization at the IETF NMRG (Network Management Research Group) as editor for intent-based networking use case documents. Through these activities, Professor Paul has made significant contributions to the development of Internet technologies worldwide as an Internet expert. Figure 3. Cloud Security System Based on the I2NSF Framework Professor Paul’s SKKU team contributed to global recognition of Korea and Sungkyunkwan University as a leading institution in Internet technology development and standardization through their IETF standardization activities on the I2NSF cloud security system and IPWAVE vehicular networking. The following shows the Internet standard documents contributed by Professor Paul in the I2NSF and IPWAVE Working Groups. I2NSF Working Group: https://datatracker.ietf.org/wg/i2nsf/documents/ IPWAVE Working Group: https://datatracker.ietf.org/wg/ipwave/documents/ Professor Paul’s research team established the I2ICF (Interface to In-Network Computing Functions) group (https://mailman3.ietf.org/mailman3/lists/i2icf.ietf.org/) at the IETF to develop standard documents for controlling and managing mobile objects connected to 5G mobile networks (e.g., software-defined vehicles, robotic cars, robots, drones). At the IETF 124 Hackathon held in Montreal, Canada in November 2025, the team demonstrated I2ICF technology as a proof of concept (POC) and is working to establish a new working group within the IETF. Figure 4 shows Professor Paul’s hackathon team, and Figure 5 shows a poster presenting the implementation and test setup of the I2ICF hackathon project. This standardization work is part of Professor Paul’s standardization tasks under the IITP project “Development of SDV Software Framework Standards for Intelligent Converged Services.” Figure 4. I2ICF Hackathon Team at IETF 124 Meeting Figure 5. I2ICF Hackathon Project Poster at IETF 124 The following shows the I2ICF drafts currently being standardized by Professor Paul’s research team. I2ICF Problem Statement: https://datatracker.ietf.org/doc/draft-jeong-opsawg-i2icf-problem-statement/ I2ICF Framework: https://datatracker.ietf.org/doc/draft-jeong-opsawg-i2icf-framework/ Professor Paul’s research team develops networking and security technologies for the Internet and actively participates in Internet standardization at the IETF as Korea’s leading standardization expert. Their research results are also published annually in top academic journals. In addition, Professor Paul serves as the publicity chair for NetSoft 2025 and program committee chair for ICMU 2025, contributing to the global recognition of Korea and Sungkyunkwan University. He currently serves as the director of the Graduate School of Convergence Security at Sungkyunkwan University, responsible for nurturing talents in convergence security.
-
- 작성일 2026-01-08
- 조회수 130
-
- [Dec 23, 2025] 2025 SKKU Rising-Fellowship: Professor Muhammad Khan of the School of Convergence Selected
- 2025 SKKU Rising-Fellowship Global Fellowship: Professor Muhammad Khan of the School of Convergence Selected 17 Professors Selected for the 2025 SKKU Rising-Fellowship This year, our university selected 17 professors as recipients of the newly established “2025 SKKU Rising-Fellowship.” The awardees are Professor Eunjin Shin (College of Social Sciences), Professor Seyoung Lee (College of Social Sciences), Professor Eunryung Lee (College of Economics), Professor Sangseok Yoo (College of Business), Professor Muhammad Khan (College of Computing and Informatics), Professor Hwaseon Lim (College of Natural Sciences), Professor Sejin Oh (College of Natural Sciences), Professor Jonghwan Ko (College of Information and Communication Engineering), Professor Sangmin Won (College of Information and Communication Engineering), Professor Jeonggyu Kim (College of Engineering), Professor Seokjun Kwon (College of Engineering), Professor Sungmin Yoon (College of Engineering), Professor Seungwon Lee (College of Medicine), Professor Mikyung Shin (Sungkyun Institute for Convergence), Professor Inki Kim (Sungkyun Institute for Convergence), Professor Wanki Bae (Institute of Nanoscience and Technology), and Professor Danbi Kang (Samsung Advanced Institute for Health Sciences and Technology). The SKKU Rising-Fellowship is an honorary title and special research support program awarded to outstanding early- to mid-career tenure-track faculty members who have either already established themselves at the highest national level or at a global standard in their academic fields, or who have demonstrated exceptional research achievements with strong potential to grow into world-class researchers. Recipients of the 2025 SKKU Rising-Fellowship were selected through a rigorous review process conducted by the selection committee, comprehensively considering the academic and qualitative excellence of research outcomes as well as global research impact. The award ceremony was held on Tuesday, December 23, with President Yoo Jibeom, representatives of the university foundation, and senior university administrators in attendance to congratulate and encourage the awardees. On behalf of the recipients, Professor Seyoung Lee of the College of Social Sciences and Professor Seokjun Kwon of the College of Engineering shared their acceptance remarks. President Yoo Jibeom stated that the university will continue to honor outstanding early- to mid-career researchers and actively support the creation of a culture and ecosystem that enable sustained challenge and growth.
-
- 작성일 2026-01-05
- 조회수 108
-
- [Dec 16, 2025] PhD student Nivedita Singh(seclab, Advisor: Hyoungshick Kim), receives KIPS Undang Student Paper Award NEW
- Nivedita Singh, a PhD student at seclab (advised by Professor Hyoungshick Kim, https://seclab.skku.edu), received the Undang Student Paper Award from the Korea Information Processing Society (KIPS) for her paper, "Behind the Screen: How Cookies Become Your Identity’s Price Tag." The study analyzed cookie and tracking behaviors across 360 e-commerce websites in 18 countries, empirically demonstrating that privacy regulations such as GDPR and CCPA are often not properly enforced. The research revealed widespread security issues, including pre-consent user tracking, cookie lifetime violations, and serious vulnerabilities like XSS and CSRF, highlighting the urgent need for improved regulatory enforcement and better cookie management practices.
-
- 작성일 2026-01-08
- 조회수 124
-
- [Dec 15, 2025] Prof. Hyungjoon Koo SecAI Lab & Prof. Sungjae Hwang SoftSec Lab, a paper accepted for NDSS '26 NEW
- Eom Jiyong, a PhD student in SecAI Lab (Advisor: Professor Hyungjoon Koo, https://secai.skku.edu/) and Omar Abusabha, a PhD student in SoftSec Lab (Advisor: Professor Sungjae Hwang, https://softsec.skku.edu/), have co-authored a paper titled "A Deep Dive into Function Inlining and its Security Implications for ML-based Binary Analysis", which has been accepted for the premier security conference, The Network and Distributed System Security Symposium 2026 (NDSS '26), and is scheduled to be presented in February 2026. Function inlining optimization is a representative technique used by compilers to improve program performance. Instead of making a function call, the compiler directly inserts the function's code at the call site, reducing the overhead associated with the function call. Function inlining is applied extensively throughout the compilation process, and some inlining occurs even when optimization options are disabled (-O0). Recently, machine learning models that assist in binary reverse engineering rely heavily on various static features of functions. However, in-depth analyses of how normal inlining optimizations affect the performance of these models have not yet been sufficiently conducted. The study shows that function inlining can significantly distort the static features used by ML models, leading to performance degradation, and that attackers can intentionally exploit this using only the compiler's default flags, without employing complex techniques such as obfuscation. To investigate this, the authors first analyzed the inlining optimization mechanisms of the LLVM compiler toolchain and systematically organized the compiler options affecting inlining. They then derived combinations of options that induce higher inlining than typical optimization levels. Subsequently, experiments were conducted across five ML-based tasks—including binary reverse engineering and malware detection—using a total of 20 ML models. Abstract: A function inlining optimization is a widely used transformation in modern compilers, which replaces a call site with the callee's body in need. While this transformation improves performance, it significantly impacts static features such as machine instructions and control flow graphs, which are crucial to binary analysis. Yet, despite its broad impact, the security impact of function inlining remains underexplored to date. In this paper, we present the first comprehensive study of function inlining through the lens of machine learning-based binary analysis. To this end, we dissect the inlining decision pipeline within the LLVM's cost model and explore the combinations of the compiler options that aggressively promote the function inlining ratio beyond standard optimization levels, which we term extreme inlining. We focus on five ML-assisted binary analysis tasks for security, using 20 unique models to systematically evaluate their robustness under extreme inlining scenarios. Our extensive experiments reveal several significant findings: i) function inlining, though a benign transformation in intent, can (in)directly affect ML model behaviors, being potentially exploited by evading discriminative or generative ML models; ii) ML models relying on static features can be highly sensitive to inlining; iii) subtle compiler settings can be leveraged to deliberately craft evasive binary variants; and iv) inlining ratios vary substantially across applications and build configurations, undermining assumptions of consistency in training and evaluation of ML models. | Professor Hyungjoon Koo | kevin.koo@skku.edu, kevinkoo001.github.io | SecAI Lab | secai.skku.edu/ | Professor Sungjae Hwang | sungjaeh@skku.edu | SoftSec Lab | softsec.skku.edu/
-
- 작성일 2026-01-08
- 조회수 137
-
- [Dec 02, 2025] SW Convergence College, Department of Real-World Media Engineering Annual Research Review Completed
- The Department of Immersive Media Engineering in the College of Software Convergence Successfully Concludes its Annual Research Review The Department of Immersive Media Engineering in the College of Software Convergence (Department Head: Professor Eun-seok Ryu) successfully held its 2025 Annual Research Review event on Thursday, November 27th at 4:30 PM in the Global R&E Lounge on the 5th floor of the International Building on the Humanities and Social Sciences Campus. This event, co-hosted by four departments—the Department of Immersive Media Engineering, the Department of Artificial Intelligence Convergence, the Department of Interaction Science, and the Department of Artificial Intelligence Convergence—featured 41 research presentations across six areas: XR/VR & Immersive Experiences, 3D Gaussian Splatting & Graphics Systems, Multimodal Understanding & Generation, Human-AI Interaction & Social Computing, AI for Emotion & Mental Health, and Data-Driven Modeling & Recommendation. The event provided a meaningful forum for examining the present and future of Immersive Media research. The presentations were presented in the form of poster exhibitions and demonstrations, and a lively discussion ensued on the practical applicability and technological scalability of the research. Preceding the event, the Industrial Advisory Board (IAB) pre-meeting was attended by representatives from major ICT and content companies and research institutes, including LG Electronics, Samsung Electronics, Sanghwa, Olympla, SOS Lab, the Electronics and Telecommunications Research Institute (ETRI), LG U+, LG HelloVision, and the Institute of Information and Communications Technology Planning and Evaluation (IITP). The participating experts exchanged in-depth views on key technology trends demanded by the industry, including XR devices, robot-based vision technology, LiDAR sensors and volumetric imaging, AI-based immersive media services, future networks, and Web3 technologies, as well as the direction of industry-academia-research cooperation. They emphasized the importance of building a collaborative research ecosystem between industry and academia. Following a presentation of research results and expert evaluations, the "Outstanding Research Award" ceremony was held, with winners selected for both the undergraduate and graduate categories. In the undergraduate category, Kang Min-gu, a student majoring in Artificial Intelligence Convergence, received the Best Research Award, while Kim Soo-hyun and Oh Kyung-jun received the Excellence in Research Award. In the graduate category, Researcher Lee Yu-bin of the Department of Artificial Intelligence Convergence received the Best Research Award. The Outstanding Research Awards went to Researcher Kim Jong-han of the Department of Realistic Media Engineering, Researcher Oh Min-woo of the Department of Metabiohealth and Researcher Park Min-soo (and his team) of the Department of Artificial Intelligence Convergence, Researcher Joo Min-jun of the Department of Realistic Media Engineering, and Researcher Jeong Ui-jun of the Department of Realistic Media Engineering. This award ceremony recognized the efforts of researchers who demonstrated creative research capabilities and the potential for practical technological advancement. Furthermore, Alumni Association Advisor Ryu Deok-hee (Honorary Chairman of Kyungdong Pharmaceutical) attended the event, offering practical advice and heartfelt encouragement to students growing into researchers who will lead future technologies. Students and researchers gained valuable insights into the practical capabilities and research attitudes required in industry, providing valuable insights into the research process. Dean Eunseok Ryu of the Department of Immersive Media Engineering, who planned the event, stated, "The Annual Research Review is a crucial forum for researchers to share their achievements and discover new collaboration opportunities. We will continue to grow as a leading global research hub in the field of Immersive Media Engineering." With support from the Ministry of Science and ICT's Virtual Convergence Graduate School project, the Department of Immersive Media Engineering operates an overseas research program and selects outstanding graduate students. It also continuously expands its research environment and international collaboration system to cultivate future talent in image processing, graphics, and artificial intelligence. This Annual Research Review concluded as a meaningful event that not only shared the achievements of undergraduate and graduate researchers, but also strengthened collaboration with industry and laid the foundation for future growth.
-
- 작성일 2025-12-09
- 조회수 341
-
- [Dec 02, 2025] Professor Woo Simon Sung-il Receives Commendation from the Minister of Science and ICT
- Professor Simon Sung-il Woo Receives Minister of Science and ICT Award Professor Simon Sung-il Woo of the Department of Software and the Graduate School of Artificial Intelligence was recognized for his world-class deepfake detection research, proactive response to solving social problems, and contributions to international cooperation and education in deepfake research. On November 12, he was selected as a person of merit for the dissemination of ICT technology achievements at the 2026 AI ICT Industry Technology Outlook Conference (https://aiictconference.kr/home/) and received the Minister of Science and ICT Award. Professor Simon Sung-il Woo and his research team are actively involved in the development of core technologies related to deepfake detection, utilizing these technologies to address social issues and collaborate internationally. Currently, he is working with the National Police Agency, the Supreme Prosecutors' Office, and the AI Safety Research Institute to develop deepfake detection technologies and cultivate AI and security talent for solving social problems and contributing to the public good.
-
- 작성일 2025-12-09
- 조회수 323
-
- [Nov 17, 2025] Professor Eun-il Park dxLab, two papers accepted for AAAI 2026 and WACV 2026
- dxLab (Advisor: Eun-il Park, https://dsl.skku.edu) has accepted two papers for publication at AAAI 2026, a prestigious conference in the field of artificial intelligence, and one paper at WACV 2026. These papers are scheduled to be presented in January (Singapore) and March (Arizona), 2026. 1) (AAAI 2026) The paper "MASP: Multi-Aspect Guided Emotion Reasoning with Soft Prompt Tuning in Vision-Language Models" was co-authored by Sang-eun Lee (Department of Artificial Intelligence Convergence, currently a researcher at ETRI) and PhD candidate Yu-bin Lee (Department of Artificial Intelligence Convergence). Professor Eun-il Park served as the corresponding author. This paper proposes MASP, a structurally integrated approach to human emotion recognition that addresses the challenges of reliance on a single image representation or limited cues, making it difficult to make detailed emotion judgments. MASP learns a Multi-Aspect Module that independently encodes six emotional cues—facial expression, scene, object, color, brightness, and action—and combines these with global image features to construct a richer visual representation. Unlike previous studies that utilized only limited perspectives, MASP separates and integrates all six cues, enabling more sophisticated emotion interpretation. Soft prompt tuning then induces an inference structure specialized for emotion recognition within the language model, achieving higher accuracy and more stable performance than previous approaches. Soft prompts enhance robustness to prompt expression changes, enabling stable inference even in real-world environments. MASP demonstrates robust performance even in challenging situations such as similar emotion classification and minority class classification, demonstrating high potential for practical applications such as human-agent interaction that require multimodal emotion understanding. 2) (WACV 2026) The paper, “Alignment and Distillation: A Robust Framework for Multimodal Domain Generalizable Human Action Recognition,” was co-authored by Jihyunbin Ji (Master's student, Department of Immersive Media Engineering) and Jooyup Lee (PhD student, Department of Artificial Intelligence Convergence), with Professor Eunil Park as the corresponding author. This paper proposes a Multimodal Alignment and Distillation for Domain Generalization (MAD-DG) framework that temporally aligns and integrates multimodal cues to address the issues of existing Human Action Recognition (HAR) models, which are limited to single modality or static fusion methods and thus vulnerable to domain changes in real-world environments. To achieve this, MAD-DG learns more robust action representations based on two key elements: First, Segment-Label Aligned Contrastive Learning (SLA-CL) compensates for the asynchronous recording issues between RGB, Optical Flow, and Audio using a temporal binding window, thereby precisely aligning semantic correspondences between modalities. This reduces noise caused by sensor delays or recording discrepancies and highlights core behavioral patterns. Furthermore, the Online Self-Distillation Temporal Module (OSDTM) constructs segment tuples of various lengths, considering that actions unfold temporally in multiple stages, and selects important combinations using soft attention. Teacher-student self-distillation creates a temporal representation that remains robust to domain changes. MAD-DG combines modality alignment with multi-scale temporal reasoning to achieve high performance in multi-source domain generalization and context-free environments (Mimetics), where existing models struggle. In particular, it actively utilizes optical flow information to construct realistic action-centric representations, demonstrating high applicability in diverse real-world environments such as complex action understanding and human-agent collaboration systems. | Professor Eun-il Park | eunilpark@skku.edu, sites.google.com/view/eunil | dxLab | sites.google.com/view/dxlab/
-
- 작성일 2025-12-09
- 조회수 354
-
- [Nov 18, 2025] Interview with the 2025 SKKU Teaching Award Winner: Professor Chang-Jun Lee
- "Don't be afraid of unfamiliar experiences." Since 2011, our university has annually presented the "SKKU Teaching Award" to faculty who have provided particularly meaningful learning experiences to students. Selection categories are divided based on teaching formats, including flipped learning, medium- and large-scale online learning, and undergraduate research credit programs. Both full-time and part-time faculty are eligible to receive the award. A total of 19 faculty members received the 2025 SKKU Teaching Award, and the awards were presented at the university's 627th anniversary celebration on September 24th. This article introduces Professor Lee Chang-jun of the Department of Global Convergence Studies, who received the 2025 SKKU Teaching Award in the undergraduate research credit program category. In an interview with Sungkyunkwan Webzine, Professor Lee shared a wide-ranging and in-depth account of his experiences and goals as a researcher, his values as an educator, and the advice he offers as an adult. Let's delve into his story through the full interview below. | Before we begin the interview, I'd like to offer my sincere congratulations on being selected for the 2025 SKKU Teaching Award. Could you share your thoughts? Sungkyunkwan University is home to some of the best researchers in their respective fields, so my expectations and aspirations were even greater when I was nominated for the Teaching Excellence Award. Indeed, I'm deeply honored and delighted to receive this award. | First, I'd like to hear about your background as a researcher. Your research has primarily focused on the interaction between media and users. Was there a particular moment that sparked your interest in this field? Actually, I didn't initially set out to research "media and user interaction." However, both academically and personally, I've always been curious about how people react in new media environments and how those experiences shape their lives and society. Since my undergraduate years, I've experienced firsthand the evolution of media, from broadcasting to the internet and then mobile. This experience has made me realize that technology isn't just a means of communication; it has the power to transform users' emotions, behavior, and even social relationships. I think this experience naturally led to my research interests. Ultimately, my research began with a "human-centered" approach, and my desire to understand the relationship between media and users through their interactions has persisted to this day. | On your current lab's website, you describe yourself as a "computational social scientist." Why do you choose this approach? I describe myself as a "computational social scientist" because I believe it best captures the starting point and methodology of my research. While my initial interest in social phenomena and human behavior began with a fundamental interest in these areas, my approach to data exploration has evolved. Now, beyond simple surveys and statistical analysis, I actively utilize large-scale digital trace data, algorithms, simulations, and artificial intelligence techniques. In other words, while my research topics maintain a focus on social science issues, my methodology combines the tools of computer science and data science. In this sense, I believe the identity of a "computational social scientist" most accurately describes my research. | I'm curious about your dreams as a researcher. My dreams as a researcher can be broadly categorized into two categories. One is an academic dream, and the other is a social one. Academically, I want to continue conducting meaningful, human-centered research in a rapidly changing media and technological landscape. In an era increasingly dominated by data and algorithms, I believe it's crucial to research without losing sight of human experiences and social context. Socially, I hope my research doesn't remain confined to academia, but can contribute to solving real-world problems, policies, industry, and even everyday life. I find the most rewarding experience when I collaborate with students on new research and when the results lead to small changes in society. Ultimately, my dream is to leave a positive mark on people and society through excellent research and education.
-
- 작성일 2025-12-09
- 조회수 333
