• Proceedings of the Korea Information Processing Society Conference
• /
• 2023.11a
• /
• pp.1100-1101
• /
• 2023

얼굴인식 모델이 서양인 얼굴에 맞춰져 있어 한국인 얼굴에 대한 인식 성능 향상이 필요하다. 본 논문에서는 얼굴인식 모델에 AIHub에서 제공하는 한국인 얼굴 데이터 셋을 추가하고, 서양인 비교되는 한국인의 특징을 추가하여 얼굴인식을 진행하였다. contrastive learning의 image pair 쌍의 적합한 비율 평가를 계층적으로 진행하여 한국인 인식 성능을 높인 VGG-Kface를 제안한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2007.11a
• /
• pp.525-528
• /
• 2007

추천시스템은 일반적으로 협동적 필터링이라는 정보 필터링 기술을 사용한다. 협동적 필터링은 유사한 성향을 갖는 다른 고객들이 상품에 대해서 매긴 평가에 기반하기 때문에 고객에게 가장 적합한 유사 이웃들을 적절히 선정해 내는 것이 추천시스템의 예측의 질 향상을 위해서 필요하다. 본 논문에서는 속성 정보를 기반으로 한 다단계 클러스터링을 통한 이웃선정 방법을 제안한다. 이 방법은 대규모 데이터 셋에서 탐색 공간을 줄이기 위해 클러스터링을 수행하여 적절한 이웃 고객들의 집합을 추출한다. 이 때, 속성 정보에 따라 단계적으로 클러스터링을 수행함으로써 보다 정제된 고객집합을 구성할 수 있도록 한다. 본 논문에서는 고객 선호도와 위치 정보를 대표적인 속성 정보로 사용함으로써 모바일 환경에서 보다 정확한 추천이 이루어질 수 있도록 한다.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2024.05a
• /
• pp.709-710
• /
• 2024

Super resolution에서 High-frequency Details를 개선하는 것이 최근 문제이다. 기존에는 Super resolution을 Regression task로 접근하므로써 L2 Loss를 사용하여 이미지가 흐릿하게 되었다. 이를 해결하기위해, Classification task로 바꾸므로써 Cross Entropy Loss을 적용하여 Cross-entropy Super-resolution (CS)를 설계한다. CS를 통해 선명도와 Details이 개선되지만, 저주파의 CE Loss 학습으로인한 Black Artifacts가 발생한다. 그래서, L2 Loss는 저주파와 같이 큰 신호에 더 초점을 맞추므로, 성능 개선을 위해 저주파를 L2 Loss에서, 고주파를 CE Loss에서 학습시킨 Frequency-specific Cross-entropy Super-resolution (FCS)을 제안한다. 우리는 왜곡에 강하며 Human의 인식과 유사한 측정지표인 Learned Perceptual Image Patch Similarity (LPIPS)로 평가한다. 실험한 모든 데이터 셋에서 우리의 FCS는 Baseline보다 LPIPS가 약 1.7배 정도 개선되었다.

• Journal of the Korea Computer Graphics Society
• /
• v.28 no.2
• /
• pp.11-19
• /
• 2022

Recently, various researches on medical image generation have been suggested, and it becomes crucial to accurately evaluate the quality and diversity of the generated medical images. For this purpose, the expert's visual turing test, feature distribution visualization, and quantitative evaluation through IS and FID are evaluated. However, there are few methods for quantitatively evaluating medical images in terms of fidelity and diversity. In this paper, images are generated by learning a chest CT dataset of non-small cell lung cancer patients through DCGAN and PGGAN generative models, and the performance of the two generative models are evaluated in terms of fidelity and diversity. The performance is quantitatively evaluated through IS and FID, which are one-dimensional score-based evaluation methods, and Precision and Recall, Improved Precision and Recall, which are two-dimensional score-based evaluation methods, and the characteristics and limitations of each evaluation method are also analyzed in medical imaging.

• Smart Media Journal
• /
• v.12 no.1
• /
• pp.24-31
• /
• 2023

Meta-learning is metacognition that instantly distinguishes between knowing and unknown. It is a learning method that adapts and solves new problems by self-learning with a small amount of data.A few-shot learning method is a type of meta-learning method that accurately predicts query data even with a very small support set. In this study, we propose a method to solve the limitations of the prototype created with the mean-point vector of each class. For this purpose, we use the few-shot learning method that created the prototype used in the few-shot learning method as the median prototype. For quantitative evaluation, a handwriting recognition dataset and mini-Imagenet dataset were used and compared with the existing method. Through the experimental results, it was confirmed that the performance was improved compared to the existing method.

• Journal of KIISE
• /
• v.44 no.9
• /
• pp.954-965
• /
• 2017

A stable power supply is very important for the maintenance and operation of the power infrastructure. Accurate power consumption prediction is therefore needed. In particular, a university campus is an institution with one of the highest power consumptions and tends to have a wide variation of electrical load depending on time and environment. For this reason, a model that can accurately predict power consumption is required for the effective operation of the power system. The disadvantage of the existing time series prediction technique is that the prediction performance is greatly degraded because the width of the prediction interval increases as the difference between the learning time and the prediction time increases. In this paper, we first classify power data with similar time series patterns considering the date, day of the week, holiday, and semester. Next, each ARIMA model is constructed based on the classified data set and a daily power consumption forecasting method of the university campus is proposed through the time series cross-validation of the predicted time. In order to evaluate the accuracy of the prediction, we confirmed the validity of the proposed method by applying performance indicators.

• Journal of the Korea Convergence Society
• /
• v.9 no.11
• /
• pp.15-21
• /
• 2018

In this study, we try to detect anomalies on the network intrusion detection system by learning only one class. We use KDD CUP 1999 dataset, an intrusion detection dataset, which is used to evaluate classification performance. One class classification is one of unsupervised learning methods that classifies attack class by learning only normal class. When using unsupervised learning, it difficult to achieve relatively high classification efficiency because it does not use negative instances for learning. However, unsupervised learning has the advantage for classifying unlabeled data. In this study, we use one class classifiers based on support vector machines and density estimation to detect new unknown attacks. The test using the classifier based on density estimation has shown relatively better performance and has a detection rate of about 96% while maintaining a low FPR for the new attacks.

• Journal of the Korea Society of Computer and Information
• /
• v.28 no.1
• /
• pp.39-47
• /
• 2023

In this paper, we propose a method to measure the similarity between golf swings in videos. As it is known that deep learning-based artificial intelligence technology is effective in the field of computer vision, attempts to utilize artificial intelligence in video-based sports data analysis are increasing. In this study, the joint coordinates of a person in a golf swing video were obtained using a deep learning-based pose estimation model, and based on this, the similarity of each swing segment was measured. For the evaluation of the proposed method, driver swing videos from the GolfDB dataset were used. As a result of measuring swing similarity by pairing swing videos of a total of 36 players, 26 players evaluated that their other swing sequence was the most similar, and the average ranking of similarity was confirmed to be about 5th. This ensured that the similarity could be measured in detail even when the motion was performed similarly.

• The Journal of the Korea institute of electronic communication sciences
• /
• v.9 no.1
• /
• pp.61-66
• /
• 2014

As a key technology of U-health, RFID can be applied to the hospitals in a variety of cases such as patient tracking, medical instrument management, and so on. However, adoption of RFID in healthcare does not reach expectations because of huge cost. Exact estimation of cost and effectiveness will boost adoption of RFID in healthcare. This study proposes a novel simulation technique to evaluate cost and effectiveness of RFID in hospital environment. To do this, this study proposes a technique for modeling patients' movements in a hospital. Based on the model, this study provides how to obtain tag event dataset by means of simulating identifications of RFID tags that are attached to patients.

• Journal of KIISE
• /
• v.42 no.6
• /
• pp.764-768
• /
• 2015

Requirement of effective image handling methods such as image retrieval has been increasing with the rising production and consumption of multimedia data. In this paper, a method of constructing more effective descriptor is proposed for robust keypoint based image retrieval. The proposed method uses information embedded in the first order and second order derivative images, in addition to the scale space image, for the descriptor construction. The performance of multi-image descriptor is evaluated in terms of the similarities in keypoints with a public domain image database that contains various image transformations. The proposed descriptor shows significant improvement in keypoint matching with minor increase of the length.