• 제목/요약/키워드: Reference dataset

검색결과 120건 처리시간 0.022초

딥페이크 영상 학습을 위한 데이터셋 평가기준 개발 (Development of Dataset Evaluation Criteria for Learning Deepfake Video)

  • 김량형;김태구
    • 산업경영시스템학회지
    • /
    • 제44권4호
    • /
    • pp.193-207
    • /
    • 2021
  • As Deepfakes phenomenon is spreading worldwide mainly through videos in web platforms and it is urgent to address the issue on time. More recently, researchers have extensively discussed deepfake video datasets. However, it has been pointed out that the existing Deepfake datasets do not properly reflect the potential threat and realism due to various limitations. Although there is a need for research that establishes an agreed-upon concept for high-quality datasets or suggests evaluation criterion, there are still handful studies which examined it to-date. Therefore, this study focused on the development of the evaluation criterion for the Deepfake video dataset. In this study, the fitness of the Deepfake dataset was presented and evaluation criterions were derived through the review of previous studies. AHP structuralization and analysis were performed to advance the evaluation criterion. The results showed that Facial Expression, Validation, and Data Characteristics are important determinants of data quality. This is interpreted as a result that reflects the importance of minimizing defects and presenting results based on scientific methods when evaluating quality. This study has implications in that it suggests the fitness and evaluation criterion of the Deepfake dataset. Since the evaluation criterion presented in this study was derived based on the items considered in previous studies, it is thought that all evaluation criterions will be effective for quality improvement. It is also expected to be used as criteria for selecting an appropriate deefake dataset or as a reference for designing a Deepfake data benchmark. This study could not apply the presented evaluation criterion to existing Deepfake datasets. In future research, the proposed evaluation criterion will be applied to existing datasets to evaluate the strengths and weaknesses of each dataset, and to consider what implications there will be when used in Deepfake research.

Land Cover Classification Map of Northeast Asia Using GOCI Data

  • Son, Sanghun;Kim, Jinsoo
    • 대한원격탐사학회지
    • /
    • 제35권1호
    • /
    • pp.83-92
    • /
    • 2019
  • Land cover (LC) is an important factor in socioeconomic and environmental studies. According to various studies, a number of LC maps, including global land cover (GLC) datasets, are made using polar orbit satellite data. Due to the insufficiencies of reference datasets in Northeast Asia, several LC maps display discrepancies in that region. In this paper, we performed a feasibility assessment of LC mapping using Geostationary Ocean Color Imager (GOCI) data over Northeast Asia. To produce the LC map, the GOCI normalized difference vegetation index (NDVI) was used as an input dataset and a level-2 LC map of South Korea was used as a reference dataset to evaluate the LC map. In this paper, 7 LC types(urban, croplands, forest, grasslands, wetlands, barren, and water) were defined to reflect Northeast Asian LC. The LC map was produced via principal component analysis (PCA) with K-means clustering, and a sensitivity analysis was performed. The overall accuracy was calculated to be 77.94%. Furthermore, to assess the accuracy of the LC map not only in South Korea but also in Northeast Asia, 6 GLC datasets (IGBP, UMD, GLC2000, GlobCover2009, MCD12Q1, GlobeLand30) were used as comparison datasets. The accuracy scores for the 6 GLC datasets were calculated to be 59.41%, 56.82%, 60.97%, 51.71%, 70.24%, and 72.80%, respectively. Therefore, the first attempt to produce the LC map using geostationary satellite data is considered to be acceptable.

Detecting the Influential Observation Using Intrinsic Bayes Factors

  • Chung, Younshik
    • Journal of the Korean Statistical Society
    • /
    • 제29권1호
    • /
    • pp.81-94
    • /
    • 2000
  • For the balanced variance component model, sometimes intraclass correlation coefficient is of interest. If there is little information about the parameter, then the reference prior(Berger and Bernardo, 1992) is widely used. Pettit nd Young(1990) considered a measrue of the effect of a single observation on a logarithmic Bayes factor. However, under such a reference prior, the Bayes factor depends on the ratio of unspecified constants. In order to discard this problem, influence diagnostic measures using the intrinsic Bayes factor(Berger and Pericchi, 1996) is presented. Finally, one simulated dataset is provided which illustrates the methodology with appropriate simulation based computational formulas. In order to overcome the difficult Bayesian computation, MCMC methods, such as Gibbs sampler(Gelfand and Smith, 1990) and Metropolis algorithm, are empolyed.

  • PDF

Fast and Accurate Visual Place Recognition Using Street-View Images

  • Lee, Keundong;Lee, Seungjae;Jung, Won Jo;Kim, Kee Tae
    • ETRI Journal
    • /
    • 제39권1호
    • /
    • pp.97-107
    • /
    • 2017
  • A fast and accurate building-level visual place recognition method built on an image-retrieval scheme using street-view images is proposed. Reference images generated from street-view images usually depict multiple buildings and confusing regions, such as roads, sky, and vehicles, which degrades retrieval accuracy and causes matching ambiguity. The proposed practical database refinement method uses informative reference image and keypoint selection. For database refinement, the method uses a spatial layout of the buildings in the reference image, specifically a building-identification mask image, which is obtained from a prebuilt three-dimensional model of the site. A global-positioning-system-aware retrieval structure is incorporated in it. To evaluate the method, we constructed a dataset over an area of $0.26km^2$. It was comprised of 38,700 reference images and corresponding building-identification mask images. The proposed method removed 25% of the database images using informative reference image selection. It achieved 85.6% recall of the top five candidates in 1.25 s of full processing. The method thus achieved high accuracy at a low computational complexity.

Sensitivity analysis of the influencing factors of slope stability based on LS-SVM

  • Xu, Juncai;Ren, Qingwen;Shen, Zhenzhong
    • Geomechanics and Engineering
    • /
    • 제13권3호
    • /
    • pp.447-458
    • /
    • 2017
  • This study proposes a sensitivity analysis method for slope stability based on the least squares support vector machine (LS-SVM) to examine the influencing factors of slope stability. The method uses LS-SVM as an algorithm for machine learning. An appropriate training dataset is established according to the slope characteristics, and a testing dataset is designed orthogonally. Results of the testing data in the experiment design are calculated after training using the LS-SVM model. The sensitivity of the slope stability of each factor is examined via gray correlation analysis. The results are consistent with those of the traditional Bishop analysis and can be used as a reference for optimizing slope design.

이미지 조작 탐지를 위한 포렌식 방법론 (A Forensic Methodology for Detecting Image Manipulations)

  • 이지원;전승제;박윤지;정재현;정두원
    • 정보보호학회논문지
    • /
    • 제33권4호
    • /
    • pp.671-685
    • /
    • 2023
  • 인공지능이 이미지 편집 기술에 적용되어 조작 흔적이 거의 없는 고품질 이미지를 생성할 수 있게 되었다. 그러나 이러한 기술들은 거짓 정보 유포, 증거 인멸, 사실 부인 등의 범죄 행위에 악용될 수 있기 때문에 이에 대응하기 위한 방안이 필요하다. 본 연구에서는 이미지 조작을 탐지하기 위해 이미지 파일 분석과 모바일 포렌식 아티팩트 분석을 수행한다. 이미지 파일 분석은 조작된 이미지의 메타데이터를 파싱하여 Reference DB와 비교분석을 통해 조작여부를 탐지하는 방법이다. Reference DB는 이미지의 메타데이터에 남는 조작 관련 아티팩트를 수집하는 데이터베이스로서, 이미지 조작을 탐지하는 기준이 된다. 모바일 포렌식 아티팩트 분석은 이미지 편집 도구와관련된 패키지를 추출하고 분석하여 이미지 조작을 탐지하도록 한다. 본 연구에서 제안하는 방법론은 기존의 그래픽적 특징기반 분석의 한계를 보완하고, 이미지 처리 기법과 조합하여 오탐을 줄일 수 있도록 한다. 연구 결과는 이러한 방법론이 디지털 포렌식 조사 및 분석에 유의미하게 활용될 수 있음을 보여준다. 또한, 조작된 이미지 데이터셋과 함께 이미지 메타데이터 파싱 코드와 Reference DB를 제공하여 관련 연구에 기여하고자 한다.

배경자료를 이용한 나무구조의 군집분석 (Tree Based Cluster Analysis Using Reference Data)

  • 최대우;구자용;최용석
    • 응용통계연구
    • /
    • 제17권3호
    • /
    • pp.535-545
    • /
    • 2004
  • 이 논문에서 제안하는 군집분석방법은 분석자료와 동일한 구조의 배경자료를 생성하고 이를 나무모형의 분류기법을 이용하여 분리해 냄으로써 변수들의 규칙으로 정의되는 군집을 형성한다. 배경자료는 reverse-arcing 알고리즘을 통하여 분석자료와 공간상에서 대비되도록 생성되며 군집이 효과적으로 식별되도록 돕는다. 이 방법은 분석자료에 이산형 변수가 혼합된 경우에도 적용할 수 있으며 모의실험자료와 실제 자료를 이용하여 제안된 알고리즘의 성능을 규명하였다.

Topographic Information Extraction from Kompsat Satellite Stereo Data Using SGM

  • Jang, Yeong Jae;Lee, Jae Wang;Oh, Jae Hong
    • 한국측량학회지
    • /
    • 제37권5호
    • /
    • pp.315-322
    • /
    • 2019
  • DSM (Digital Surface Model) is a digital representation of ground surface topography or terrain that is widely used for hydrology, slope analysis, and urban planning. Aerial photogrammetry and LiDAR (Light Detection And Ranging) are main technology for urban DSM generation but high-resolution satellite imagery is the only ingredient for remote inaccessible areas. Traditional automated DSM generation method is based on correlation-based methods but recent study shows that a modern pixelwise image matching method, SGM (Semi-Global Matching) can be an alternative. Therefore this study investigated the application of SGM for Kompsat satellite data of KARI (Korea Aerospace Research Institute). Firstly, the sensor modeling was carried out for precise ground-to-image computation, followed by the epipolar image resampling for efficient stereo processing. Secondly, SGM was applied using different parameterizations. The generated DSM was evaluated with a reference DSM generated by the first pulse returns of the LIDAR reference dataset.

Korean-specific iodine S values for use in internal dosimetry

  • Tae-Eun Kwon;Yoonsun Chung;Choonsik Lee
    • Nuclear Engineering and Technology
    • /
    • 제55권12호
    • /
    • pp.4659-4663
    • /
    • 2023
  • The use of iodine S values derived using the International Commission Radiological Protection (ICRP) phantoms may introduce significant bias in internal dosimetry for Koreans due to anatomical variability. In the current study, we produced an extensive dataset of Korean S values for selected five iodine radioisotopes (I-125, I-129, I131, I-133, and I-134) for use in radiation protection. To calculate S values, we implemented Monte Carlo simulations using the Mesh-type Reference Korean Phantoms (MRKPs), developed in a high-quality/fidelity mesh format. Noticeable differences were observed in S value comparisons between the Korean and ICRP reference phantoms with ratios (Korean/ICRP) widely ranging from 0.16 to 6.2. The majority of S value ratios were lower than the unity in Korean phantoms (interquartile range = 0.47-1.28; mean = 0.96; median = 0.69). The S values provided in the current study will be extensively utilized in iodine internal dosimetry for Koreans.

3D 인쇄방법으로 제작된 치과용 다이 모델의 정확도 평가연구 (A study on the accuracy evaluation of dental die models manufactured by 3D printing method)

  • 장연
    • 대한치과기공학회지
    • /
    • 제41권4호
    • /
    • pp.287-293
    • /
    • 2019
  • Purpose: To evaluate the accuracy of the 3D printed die models and to investigate its clinical applicability. Methods: Stone die models were fabricated from conventional impressions(stone die model; SDM, n=7). 3D virtual models obtained from the digital impressions were manufactured as a 3D printed die models using a 3D printer(3D printed die models;3DM, n=7). Reference model, stone die models and 3D printed die models were scanned with a reference scanner. All dies model dataset were superimposed with the reference model file by the "Best fit alignment" method using 3D analysis software. Statistical analysis was performed using the independent t-test and 2-way ANOVA (α=.05). Results: The RMS value of the 3D printed die model was significantly larger than the RMS value of the stone die model (P<.001). As a result of 2-way ANOVA, significant differences were found between the model group (P<.001) and the part (P<.001), and their interaction effects (P<.001). Conclusion: The 3D printed die model showed lower accuracy than the stone die model. Therefore, it is necessary to further improve the performance of 3D printer in order to apply the 3D printed model in prosthodontics.