-
- [SKKU's Special lecture] YouTube comment event, meal with Professor Seo Eui-sung
- In June, undergraduate Yoo Gun-wook, won the commentary event for the Earth's Month in the special lecture of SKKU and used a meal voucher with Professor Seo Eui-sung :) Q: Going one step further, what were the efforts to prevent climate change using AI? A: Hello, it's interesting that efforts to reduce carbon emissions lead to lower power costs, which in turn is in line with the efficiency that engineers should pursue. I know that various and creative attempts are being made to reduce the power consumption of data centers overseas, and it is surprising that domestic data centers are also developed.. I'm an undergraduate who wants to work in cloud computing, and I find in my lab research that this way can also improve the cost problem. It seems to be an era when algorithms that train models efficiently at low cost or scheduling techniques are becoming important to use low-carbon energy. I think that the rhetoric that the speed of AI development should be slowed down may be cited as the reason for carbon emission later on. Thank you so much for the good lecture Corresponding link ↓ [Sung University's grand lecture] AI, Cloud, and Effects on Climate Warming | Seo Eui-sung, Professor of Software at Sungkyunkwan University - YouTube
-
- 작성일 2023-08-21
- 조회수 274
-
- [Software-centered university council] 2023 Joint Hackathon Competition
-
- 작성일 2023-07-10
- 조회수 261
-
-
- [Promotion] Finding a college - Lee Eun-seok, dean of Software Convergence University (electronic 81)
-
In the 10th order of the planning tour
, I visited a software convergence university that encompasses natural science campus and humanities and social science campus. Software Convergence University is a new university that was newly created as a software convergence university in 2021 by combining existing software universities and global convergence departments. It is a university that studies cutting-edge studies based on various SW and AI, including undergraduate and graduate schools. Software Convergence University has produced a total of 1,547 alumni, including 819 bachelor's degrees and 728 master's and doctorate degrees, and operates a "global AI cluster" that integrates AI graduate school, AI convergence department, and AI system engineering as a university that studies and fosters talent. Lee Eun-seok, president of Software Convergence University, said, "Our university's biggest goal is to cultivate SW-AI professionals with global capabilities," adding, "We will create a heartwarming engineer who can impress the community by allowing high-quality education and the latest research in the best environment. Q : Please introduce Software Convergence University.Software Convergence University A : (from here referred to as SCU) is the newest young organization among the 16 colleges in the school. Currently, there are 45 full-time professors and 69 non-full-time professors, and despite their academic advancement, we can probably introduce Sugichiin, our university's founding ideology, as one of the best practices. Each member is working hard to hone himself and devote himself to the community for the future for mankind. In addition to the common graduation requirements required by the school, students must meet all stricter graduation requirements (internships, industry-academic cooperation projects, 10,000-line projects, graduation works, etc.). Through this, we strive to acquire the basic skills and competitiveness required by the field as a software engineer, and we require SW mentoring for elementary, middle, and high school students across the country, as well as other universities and silver generations (Sung Kyun SW mentoring). In addition, we are trying to run an ICT overseas volunteer group for countries in a more difficult environment so that we can share the benefits of our lives across borders. It is an educational ideology and talent shared with our Soyong University to cultivate true talent who is not ashamed to talk about humanity or the future, not to get a job at a good company or university. Q : I'm curious about the history of Software Convergence University A : Founded with the vision of fostering SW-AI professionals with global capabilities,our university was expanded to Software Convergence University in 2015 and 2021 by integrating software and computer engineering departments and AI and data science professors and organizations. Specifically, it is an undergraduate organization consisting of software and global convergence departments, and a large organization consisting of 2,600 undergraduates, 800 graduate students, and 120 faculty members in 14 graduate programs. From 2024, the Department of Intelligent SW, an employment contract-type department jointly operated with Samsung Electronics, will begin both academic and master's courses. For reference, there are 10 industry-academic cooperation talent training courses (including employment contract types) supported by Samsung Electronics nationwide, including semiconductors and telecommunications, and SW-AI-related programs are owned only by our universities. Q : What are the strengths and characteristics of our university's software convergence university A : The main source of funding for running University may be tuition, but especially because of the academic nature that it is absolutely necessary to have the latest equipment and environment, it is necessary to reduce dependence on tuition in the school area and have the ability to grow on its own. In that respect, our SCU has the efforts and capabilities to secure resources for voluntary manpower training. It is currently in progress and has secured 62 billion won in total through artificial intelligence graduate school projects, software-oriented university projects, and luxury talent training projects, and if you add an intelligent SW department, it will reach up to 100 billion won. Through this, the educational and research environment of undergraduate and graduate schools is advanced. And in the current era of digital transformation, it is serving as an advancement in SW-AI education for all students to improve SW-based convergence skills as well as major skills of students throughout the university. Moreover, AI education for non-technical professors already has more than 200 people, and convergence and inclusion among actual university members have become possible through joint project discovery. In summary, it is recognized for its strengths internally as the core of convergence and externally as a competitive subject of education and research externally. Q : What is the research area or research topic or project that Software Convergence University focuses on A : I think that various fields must develop evenly to effectively cope with rapid changes in society and technology and fulfill the university's original role. However, in order to respond to certain urgent demands like these days and to provide the necessary manpower and research results in a timely manner, we are currently gathering training and research personnel to AI. AI-related education systems include AI graduate school, AI convergence major, AI convergence department, AI system engineering department, intelligent SW department, and smart factory department. This is also an important infrastructure for conducting AI-related research. To revitalize AI research, 26 global AI institutions are grouped to operate a 'Global AI Cluster'. I think this laid the foundation for substantial AI education and research. In addition to this, our university has competitive edge in system SW, SW security, software engineering, etc. Q : If there is a development goal that Software Convergence University aims for A : Under the big goal of fostering SW-AI professionals with global capabilities, we would like to create a heartwarming engineer who can always contribute to the community and impress the community. Graduate schools will have advanced research competitiveness to produce high-quality papers, and undergraduate departments will focus on connecting with industry, reflecting the requirements and technical trends of the field, and will be able to acquire the work skills needed in the field. To this end, we operate a curriculum innovation committee composed of industry experts and make efforts to reflect the opinions of the field. Q. What do you want to say to the 250,000 SKKU alumni? In order to become a prestigious university A, it is necessary for universities, foundations, and alumni associations to do their best as a Trinity. Among them, the alumni association has served as a strong support for university members. Although there is a lot of interest and donations for alma mater and juniors, it is time to think about how universities can contribute to the alumni organization rather than such a one-way communication support system. In particular, as our university deals with SW-AI technology, which is the basis of all industries, I think it can serve as a technical backstop for alumni companies. If you need help, please contact us anytime and build a mutually beneficial and constructive relationship. -
- 작성일 2023-07-05
- 조회수 346
-
- [Research News] Security Engineering Laboratory has been approved for publication in The Web Conference (WWW) 2023 paper
- Security Engineering Laboratory has been approved for publication in The Web Conference (WWW) 2023 paper. The paper "AppSniffer: Towards Robust Mobile App Fingerprinting Against VPN" by Sanghak Oh and Professor Hyoungshick Kim of the Security Engineering Laboratory has been approved for publication in the Web Conference (WWW) 2023 (https://www2023.thewebconf.org) (BKIF=4). It will also be announced in Texas in the U.S. in April 2023. This paper presents the limitation that existing mobile app fingerprinting systems can be easily bypassed through VPN technology through experiments, and proposes AppSniffer, a new mobile app fingerprinting system, to fix this error. AppSniffer is designed to analyze mobile app traffic to extract feature points, even if it is generated in a VPN environment, and finally perform mobile app fingerprinting through ensemble modeling. In this paper, 100 mobile app traffic was collected in both general and VPN environments, and experiments showed that AppSniffer can perform mobile app fingerprinting in all environments (general/VPN environments), showing robustness to VPN traffic. [Paper information] Sanghak Oh, Minwook Lee, Hyunwoo Lee, Elisa Bertino, and Hyoungshick Kim. AppSniffer: Towards Robust Mobile App Fingerprinting Against VPN” In Proceedings of the ACM 32nd Web Conference: WWW 2023, Austin, USA, 2023 Abstract: Application fingerprinting is a useful data analysis technique for network administrators, marketing agencies, and security analysts. For example, an administrator can adopt application fingerprinting techniques to determine whether a user's network access is allowed. Several mobile application fingerprinting techniques (e.g., Flowprint, AppScanner, and ET-BERT) were recently introduced to identify applications using the characteristics of network traffic. However, we find that the performance of the existing mobile application fingerprinting systems significantly degrades when a virtual private network (VPN) is used. To address such a shortcoming, we propose a framework dubbed AppSniffer that uses a two-stage classification process for mobile app fingerprinting. In the first stage, we distinguish VPN traffic from normal traffic; in the second stage, we use the optimal model for each traffic type. Specifically, we propose a stacked ensemble model using Light Gradient Boosting Machine (LightGBM) and a FastAI library-based neural network model to identify applications' traffic when a VPN is used. To show the feasibility of AppSniffer, we evaluate the detection accuracy of AppSniffer for 100 popularly used Android apps. Our experimental results show that AppSniffer effectively identifies mobile applications over VPNs with F1-scores between 80.71% and 92.66% across four different VPN protocols. In contrast, the best state-of-the-art method (i.e., AppScanner) demonstrates significantly lower F1-scores between 31.69% and 48.22% in the same settings. Overall, when normal traffic and VPN traffic are mixed, AppSniffer achieves an F1-score of 88.52%, which is significantly better than AppScanner that shows an F1-score of 73.93%. 김형식 | hyoung@skku.edu | Security Engineering Lab. | http://seclab.skku.edu/
-
- 작성일 2023-02-01
- 조회수 485
-
- [Research News] Professor Heo Jae-Pil's lab was approved for publishing two papers in AAAI 2023.
- Two papers from the Visual Computing Laboratory (Professor Heo Jae-pil) have been approved for publication at the AAAI Conference on Artificial Intelligence 2023 (AAAI-23), a top-tier academic conference in the field of artificial intelligence. Paper #1: "Minority-Oriented Vicinity Expansion with Attentive Aggregation for Video Long-Tailed Recognition" WonJun Moon, Hyun Seok Seong, and Jae-Pil Heo Paper #2: “Progressive Few-shot Adaption of Generative Model with Align-free Spatial Correlation” Jongbo Moon*, Hyunjun Kim*, and Jae-Pil Heo (*: equal contribution) The paper "Minority-Oriented Vicinity Expansion with Attractive Aggregation for Video Long-Tested Recognition" deals with the issue of data imbalance that occurs when video data is acquired. First of all, we raise additional issues to consider in the video sector, along with data imbalances: 1) weak supervision of video data and 2) the size of traditional video data makes the pre-trained network unsuitable for downstream operations. To fix this error, the paper introduces two Attractive Aggregators and proposes a modified extrapolation and interpolation technique that increases the diversity of classes with fewer data to solve the data imbalance problem. Experiments have shown that the proposed method has resulted in consistent performance improvements in benchmarks. In addition, experiments have confirmed the importance of two issues that we have argued should be addressed simultaneously with data imbalance through ablation studies. The "Progressive Few-shot Adaptation of Generative Model with Align-free Spatial Correlation" paper addresses the problem of adapting the GANs model with only a very small number of target domain images. Since the use of common methods such as fine-tuning is vulnerable to mode-collapse, methods for learning that the images generated by the Source and Target models respectively maintain a relative distance have been recently studied. However, there is a problem that 1) the method of measuring the distance with the overall feature of the image loses the detailed feature of the Source model, and 2) the method of learning to maintain the consistency of the image patch unit feature loses the structural characteristic of the target domain. To fix this error, the paper aimed to adaptation that reflects the structural characteristics of the target domain through comparisons between meaningful areas (e.g., comparisons between human and character eye areas) while preserving the detailed features of the Source model. To this end, we proposed 1) Progressive Adaptation that reduces Domain Gap, 2) Align-free Spatial Correlation for comparison between meaningful areas, and 3) Importance Sampling methods. It was confirmed that the proposed method through various experiments showed excellent performance in quantitative and qualitative evaluations, especially in human evaluations. [About paper #1] Minority-Oriented Vicinity Expansion with Attentive Aggregation for Video Long-Tailed Recognition WonJun Moon, Hyun Seok Seong, and Jae-Pil Heo Thirty-Seventh AAAI Conference on Artificial Intelligence (AAAI), 2023 Abstract: A dramatic increase in real-world video volume with extremely diverse and emerging topics naturally forms a long-tailed video distribution in terms of their categories, and it spotlights the need for Video Long-Tailed Recognition (VLTR). In this work, we summarize the challenges in VLTR and explore how to overcome them. The challenges are: (1) it is impractical to re-train the whole model for high-quality features, (2) acquiring frame-wise labels requires extensive cost, and (3) long-tailed data triggers biased training. Yet, most existing works for VLTR unavoidably utilize image-level features extracted from pretrained models which are task-irrelevant, and learn by video-level labels. Therefore, to deal with such (1) task-irrelevant features and (2) video-level labels, we introduce two complementary learnable feature aggregators. Learnable layers in each aggregator are to produce task-relevant representations, and each aggregator is to assemble the snippet-wise knowledge into a video representative. Then, we propose Minority-Oriented Vicinity Expansion (MOVE) that explicitly leverages the class frequency into approximating the vicinity distributions to alleviate (3) biased training. By combining these solutions, our approach achieves state-of-the-art results on large-scale VideoLT and synthetically induced Imbalanced-MiniKinetics200. With VideoLT features from ResNet-50, it attains 18% and 58% relative improvements on head and tail classes over the previous state-of-the-art method, respectively. [About paper #2] Progressive Few-shot Adaption of Generative Model with Align-free Spatial Correlation Jongbo Moon*, Hyunjun Kim*, and Jae-Pil Heo (*: equal contribution) Thirty-Seventh AAAI Conference on Artificial Intelligence (AAAI), 2023 Abstract: In few-shot generative model adaptation, the model for target domain is prone to the mode-collapse. Recent studies attempted to mitigate the problem by matching the relationship among samples generated from the same latent codes in source and target domains. The objective is further extended to image patch-level to transfer the spatial correlation within an instance. However, the patch-level approach assumes the consistency of spatial structure between source and target domains. For example, the positions of eyes in two domains are almost identical. Thus, it can bring visual artifacts if source and target domain images are not nicely aligned. In this paper, we propose a few-shot generative model adaptation method free from such assumption, based on a motivation that generative models are progressively adapting from the source domain to the target domain. Such progressive changes allow us to identify semantically coherent image regions between instances generated by models at a neighboring training iteration to consider the spatial correlation. We also propose an importance-based patch selection strategy to reduce the complexity of patch-level correlation matching. Our method shows the state-of-the-art few-shot domain adaptation performance in the qualitative and quantitative evaluations. Heo Jae-Pil | jaepilheo@skku.edu | Visual Computing Lab | https://sites.google.com/site/vclabskku/
-
- 작성일 2023-02-01
- 조회수 499
-
- [Research News] Professor Cha Soo-young's paper was approved for publication in ICSE 2023
- A paper by Professor Cha Soo-young (co-communications) of the Department of Software has been approved for publication in the ICSE 2023 (The IEEE/ACM International Conference on Software Engineering), a top international conference in software engineering. This paper "Learning Seed-Adaptive Mutation Strategies for Greybox Fuzzing" will be published in Melbourne, Australia in May 2023. In this paper, we propose a technique called 'SEAMFUZZ', which adaptively changes the 'mutation strategy' with seed input, which has a significant impact on the performance (e.g., error detection ability) of Grey-box fuzzing. To this end, this paper proposes a 'Customized Thompson Sampling' algorithm that learns a mutation strategy optimized for each seed input based on data generated during purging. As a result, this study succeeded in detecting higher code coverage and many errors than conventional technologies in various benchmark programs. [thesis information] - “Learning Seed-Adaptive Mutation Strategies for Greybox Fuzzing” - Myungho Lee, Sooyoung Cha, and Hakjoo Oh - The IEEE/ACM International Conference on Software Engineering (ICSE 2023) - Abstract: In this paper, we present a technique for learning seed-adaptive mutation strategies for fuzzers. The performance of mutation-based fuzzers highly depends on the mutation strategy that specifies the probability distribution of selecting mutation methods. As a result, developing an effective mutation strategy has received much attention recently, and program-adaptive techniques, which observe the behavior of the target program to learn the optimized mutation strategy per program, have become a trending approach to achieve better performance. They, however, still have a major limitation; they disregard the impacts of different characteristics of seed inputs which can lead to explore deeper program locations. To address this limitation, we present SEAMFUZZ, a novel fuzzing technique that automatically captures the characteristics of individual seed inputs and applies different mutation strategies for different seed inputs. By capturing the syntactic and semantic similarities between seed inputs, SEAMFUZZ clusters them into proper groups and learns effective mutation strategies tailored for each seed cluster by using the customized Thompson sampling algorithm. Experimental results show that SEAMFUZZ improves both the path-discovering and bug-finding abilities of state-of-the-art fuzzers on real-world programs. 차수영 | sooyoung.cha@skku.edu | Software Analysis Lab | https://sal.skku.ed
-
- 작성일 2023-01-31
- 조회수 572
-
- [Promotion] Professor Muhammad Khan of the Department of Global Convergence, selected as an honorary citizen of Seoul
- Professor Muhammad Khan of the Department of Global Convergence, selected as an honorary citizen of Seoul - By Utilizing of artificial intelligence and computer science technology, developed new technologies related to civil safety. The only foreign researcher studying in Korea of the world's top 1% researcher (HCR). Professor Muhammad Khan of the Department of Global Convergence has selected as an honorary citizen of Seoul on December 9 (Fri). At the honorary citizenship award ceremony, 18 foreigners from 16 countries, including Professor Muhammad Khan, were selected as "honorary citizens of Seoul." The honorary citizenship of foreigners in Seoul began in 1958 by awarding merit citizenship to foreigners who helped rebuild the city after the war. Currently, it is given to foreigners living in Seoul, foreign heads of state, and diplomatic envoys who have contributed to the development of Seoul. According to the Seoul Metropolitan Government, 895 people from 100 countries received honorary citizenship cards from Seoul as of November 30, 2022. Professor Muhammad Khan was selected as an honorary citizen in recognition of his contribution to improving the level of science and technology by developing new technologies related to civil safety, such as analyzing fire sites and monitoring abnormal situation videos using artificial intelligence and computer technology. In addition, Professor Muhammad Khan was selected as the "World's Most Influential Top 1% Researcher (HCR)" by Clarivate. He was the only foreign researcher in Korea to be selected as the world's top 1% researcher.
-
- 작성일 2023-01-11
- 조회수 572
-
- [Promotion] 2022 SKKU Graduate Student Won the Grand Prize for Papers
- ◦ Excellence Award in Humanities and Social Sciences: Lee Hae-in/Lee Sun-hong/Jeong Hae-sun, Department of Artificial Intelligence Convergence ◦ Excellence Award in Natural Science: Lee Bo-hyun, Department of Software ◦ Excellence Award in Natural Science: Kim Ji-hwan, Department of Artificial Intelligence ◦ Encouragement Award in the field of natural science: Na Chul-won, Department of Artificial Intelligence
-
- 작성일 2023-01-03
- 조회수 494
-
- [Research News] Professor Heo Jae-Pil's lab has been approved to publish a paper in ECCV 2022
- Two papers by the Visual Computing Laboratory (Guidance Professor: Jae-Pil Heo) have been approved for publication at the European Conference on Computer Vision 2022, a top-tier academic conference in the field of computer vision and artificial intelligence. Thesis #1: "Tailoring Self-Supervision for Superimposed Learning" (Master of Artificial Intelligence WonJun Moon, Ph.D. in Artificial Intelligence, Ji-Hwan Kim) Thesis #2: "Difficulty-Aware Simulator for Open Set Recognition" (Master of Artificial Intelligence WonJun Moon, Master of Artificial Intelligence Junho Park, Master of Artificial Intelligence Hyun-Seok Seong, Master of Artificial Intelligence, and Master of Artificial Intelligence Cheol-Ho Cho) In the "Tailing Self-Supervision for Supervisory Learning" paper, we first pointed out the problems that may arise when the Self-supervision Task has been applied additionally without any special changes in the Supervisory Learning environment. When the Self-supervision Task is applied as an assistant to the Objective of Supervised Learning, we present three characteristics that Self-supervision Task should have and propose a new task called Localization Rotation that satisfies them. We find that the proposed method brings consistent performance improvements on several benchmarks that can test the robustness and generalization capabilities of the Deep Learning model. The paper "Difficulty-Aware Simulator for Open Set Recognition" presents a new method for simulating virtual samples for Open Set Recognition. The Open Set Recognition problem is a problem that identifies a new class of data that has not been experienced in learning and is an essential element of technology for applying artificial intelligence to the real world. Existing methods also generated and utilized virtual samples for model learning, but this paper confirmed that existing technologies are difficult to cope with Open Set samples of various difficulties. We propose a Difficulty-Aware Simulator framework that simulates Open Set samples of various difficulty levels. The proposed technique produced virtual samples by difficulty level from the standpoint of the classifier as intended. Using this, we achieved high performance in the Open Set Recognition field. [Thesis #1 Information] Tailoring Self-Supervision for Supervised Learning WonJun Moon, Ji-Hwan Kim, and Jae-Pil Heo European Conference on Computer Vision (ECCV), 2022 Abstract: Recently, it is shown that deploying a proper self-supervision is a prospective way to enhance the performance of supervised learning. Yet, the benefits of self-supervision are not fully exploited as previous pretext tasks are specialized for unsupervised representation learning. To this end, we begin by presenting three desirable properties for such auxiliary tasks to assist the supervised objective. First, the tasks need to guide the model to learn rich features. Second, the transformations involved in the self-supervision should not significantly alter the training distribution. Third, the tasks are preferred to be light and generic for high applicability to prior arts. Subsequently, to show how existing pretext tasks can fulfill these and be tailored for supervised learning, we propose a simple auxiliary self-supervision task, predicting localizable rotation (LoRot). Our exhaustive experiments validate the merits of LoRot as a pretext task tailored for supervised learning in terms of robustness and generalization capability. [Thesis #2 Information] Difficulty-Aware Simulator for Open Set Recognition WonJun Moon, Junho Park, Hyun Seok Seong, Cheol-Ho Cho, and Jae-Pil Heo European Conference on Computer Vision (ECCV), 2022 Abstract: Open set recognition (OSR) assumes unknown instances appear out of the blue at the inference time. The main challenge of OSR is that the response of models for unknowns is totally unpredictable. Furthermore, the diversity of open set makes it harder since instances have different difficulty levels. Therefore, we present a novel framework, DIfficulty-Aware Simulator (DIAS), that generates fakes with diverse difficulty levels to simulate the real world. We first investigate fakes from generative adversarial network (GAN) in the classifier's viewpoint and observe that these are not severely challenging. This leads us to define the criteria for difficulty by regarding samples generated with GANs having moderate-difficulty. To produce hard-difficulty examples, we introduce Copycat, imitating the behavior of the classifier. Furthermore, moderate- and easy-difficulty samples are also yielded by our modified GAN and Copycat, respectively. As a result, DIAS outperforms state-of-the-art methods with both metrics of AUROC and F-score.
-
- 작성일 2022-09-22
- 조회수 649