• Journal of the Korea Society of Computer and Information
• /
• v.28 no.10
• /
• pp.67-76
• /
• 2023

Although deep learning models are making innovative achievements in the field of computer vision, the problem of vulnerability to adversarial examples continues to be raised. Adversarial examples are attack methods that inject fine noise into images to induce misclassification, which can pose a serious threat to the application of deep learning models in the real world. In this paper, we propose a model that detects adversarial examples using differences in predictive values between edge-learned classification models and underlying classification models. The simple process of extracting the edges of the objects and reflecting them in learning can increase the robustness of the classification model, and economical and efficient detection is possible by detecting adversarial examples through differences in predictions between models. In our experiments, the general model showed accuracy of {49.9%, 29.84%, 18.46%, 4.95%, 3.36%} for adversarial examples (eps={0.02, 0.05, 0.1, 0.2, 0.3}), whereas the Canny edge model showed accuracy of {82.58%, 65.96%, 46.71%, 24.94%, 13.41%} and other edge models showed a similar level of accuracy also, indicating that the edge model was more robust against adversarial examples. In addition, adversarial example detection using differences in predictions between models revealed detection rates of {85.47%, 84.64%, 91.44%, 95.47%, and 87.61%} for each epsilon-specific adversarial example. It is expected that this study will contribute to improving the reliability of deep learning models in related research and application industries such as medical, autonomous driving, security, and national defense.

• Journal of the Korean housing association
• /
• v.18 no.3
• /
• pp.81-89
• /
• 2007

This study is aimed at understanding objectively on the visual perception of atypical space and form in contemporary architecture. It is accomplished to compose the various form and the atypical space in contemporary architecture, which by means of applying digital technology to architecture is able to materialize complicated form and space. Therefore this the sis is needed to understand user's senses and affects in atypical space and form. Atypical space is able to interpret geometrically and analyze quantitatively into visual perception. This analysis is applied to Isovist used as quantitative analysis recently, and physical analysis of space used as quantitative analysis defined to vertical elements, which is analysed comparatively to visual perception in observer's point and characteristics in spatial elements. Examples are selected to four examples represented to typical and atypical space in contemporary houses. As examples are analysed to visual perception of spatial character, this study is intended to understand to spatial character in aspects of spatial composition and observer's visual perception. In results of perceptive analysis on four spatial examples, Atypical spaces are composed of changeable spatiality, which are composed changeably of the relational method with adjacent spaces. Also, typical spaces are composed monotonously of spatiality and relational method on control of the length in boundary plane and openness.

• Journal of Information Technology Services
• /
• v.14 no.1
• /
• pp.129-143
• /
• 2015

This study has two goals. The first goal is to investigate whether worked examples are effective in the ill-defined domain with on-line learning and the second goal is to find out which components (conceptual or procedural knowledge) of worked examples are effective factor at the given learning environment. We carried out three experiments in which Korean undergraduate or graduate students were working in three or four conditions of worked examples (CWE, PWE, CPWE, or the control group). While experiment 1 conducted in on-line learning environment did not find any effect and difference among groups and also any logical reason for those results, experiment 2 conducted in completely controlled laboratory setting with less knowledgeable students showed the clear difference among groups by the order CPWE, PWE, and CWE. Experiment 3 in which highly knowledgeable and motivated students were presented the same materials in more controlled on-line learning environment indicated the difference among groups by the order CWE, CPWE, and PWE. The results were discussed within the framework of cognitive load theory.

• Journal of the Korean Institute of Telematics and Electronics A
• /
• v.32A no.11
• /
• pp.94-104
• /
• 1995

In this paper, we present a new multi-level logic synthesis method for producing the multi-level circuits which can be easily tested by the pseudoexhaustive testing techniques. The method consists of four stages. In the first stage, it generates the minimum variable supports for each output of a multiple-output function. In the second stage, it removes the minimum variable supports which if used to implement the outputs, lead to inefficient pseudoexhaustive test. In the third stage, it determines the minimum variable support and logic (uncomplementary or complementary logic) for each output. In the fourth stage, it performs the multi-level logic synthesis so that each output. In the fourth stage, it performs the multi-level logic synthesis so that each output has the minimum variable support and logic determined in the third stage. To evaluate the performance and quality of the proposed method, we have experimented on the 56 benchmark examples. The results show that for 56 examples, our method obtains better results than MIS in terms of testability. Moreover, the method produces better results for 19 examples and the same results for 12 examples compared with MIS in terms of literal count although it has been developed to improve the testability.

• 한국HCI학회:학술대회논문집
• /
• 2009.02a
• /
• pp.257-262
• /
• 2009

This paper studies an automatic expansion of dialogue examples using the logs of an example-based dialogue system. Conventional approaches to example-based dialogue system manually construct dialogue examples between humans and a Chatbot, which are labor intensive and time consuming. The proposed method automatically classifies natural utterance pairs and adds them into dialogue example database. Experimental results show that lexical, POS and modality features are useful for classifying natural utterance pairs, and prove that the dialogue examples can be automatically expanded using the logs of a dialogue system.

• Proceedings of the Korean Society for Bioinformatics Conference
• /
• 2003.10a
• /
• pp.95-100
• /
• 2003

In this paper, we propose solutions to resolve the problem of many spelling variants and the problem of lack of annotated corpus for training, which are two among the main difficulties in named entity recognition in biomedical domain. To resolve the problem of spotting valiants, we propose a use of edit-distance as a feature for SVM. And we propose a use of virtual examples to automatically expand the annotated corpus to resolve the lack-of-corpus problem. Using virtual examples, the annotated corpus can be extended in a fast, efficient and easy way. The experimental results show that the introduction of edit-distance produces some improvements in protein name recognition performance. And the model, which is trained with the corpus expanded by virtual examples, outperforms the model trained with the original corpus. According to the proposed methods, we finally achieve the performance 75.80 in F-measure(71.89% in precision,80.15% in recall) in the experiment of protein name recognition on GENIA corpus (ver.3.0).

• Journal of the Korean School Mathematics Society
• /
• v.23 no.3
• /
• pp.299-321
• /
• 2020

PISA, organized by the OECD, started with the worries about what competencies students need in preparation for a changing future society. Starting with the first main survey in 2000, PISA, which was administered every three years, is preparing for the eighth cycle. PISA 2022 is a cycle in which mathematics becomes the main domain in 10 years, and the definition of mathematics literacy, mathematical framework, and illustrative examples were released. Therefore, in this study, the definition of PISA mathematics literacy and the trends on the mathematical framework were examined, and the characteristics of the illustrative examples introduced together with the PISA 2022 mathematical framework were analyzed. Through this, implications were drawn for the successful implementation of the 2015 revised curriculum and assessment.

• Journal of Elementary Mathematics Education in Korea
• /
• v.10 no.2
• /
• pp.151-169
• /
• 2006

This study is to adjust the Theory in the Mathematics Education, apply it to learning mathematics and to analyse its effectiveness. The results of the study are summarized as follows. First, because learning mathematics is hierarchical, teachers must make and use a task analysis table classified by units. Second, development age and the retention of mathematics concepts are intimately associated with cognitive development theory. Third, learning mathematics through cognitive processes enhances a student's scholastic achievement. Fourth, students interests and self-confidence can be enhanced through the presentation of both examples and non-examples. We cannot understand the higher-order concepts of mathematics by only its definitions. The only way of understanding such concepts is to have experience through suitable examples.

• JSTS:Journal of Semiconductor Technology and Science
• /
• v.8 no.3
• /
• pp.210-220
• /
• 2008

We review three examples (an on-chip transmission line resonator [1], a phase-locked loop [2], and an analog-to-digital converter [3]) of design tradeoffs which can in fact be circumvented; the key in each case is that the parameters that seem to trade off with each other are actually separated in time or space. This paper is an attempt to present these designs in such a way that this common approach can hopefully be applied to other circuits. We note reader that this paper is not a new contribution, but a review in which we highlight the common theme from our published works [1-3]. We published a similar paper [4], which, however, used only two examples from [1] and [2]. With the newly added content from [3] in the list of our examples, the present paper offers an expanded scope.

• Journal of the Korean School Mathematics Society
• /
• v.17 no.2
• /
• pp.291-319
• /
• 2014

The purpose of this study is to investigate students' thinking and understanding through working on Worked-out Examples on numbers and operations, specifically, radical and real numbers and operations in the middle grades. For this purpose, we developed a set of Worked-out Examples; middle school students independently worked on them. Then two students were interviewed. These data were analyzed by using the framework of mathematical proficiency. The data analysis suggested that the students seemed to go through the processes involving a combination of understanding and computation, computation and reasoning, and understanding, computation and reasoning. Also, it appeared that most of the students have difficult solving problems involving with radical and real numbers in related to strategic competence.