no code implementations • 6 Mar 2024 • Sungho Kang, YeongHyeon Park, Hyunkyu Park, Juneho Yi
To address the problem of scene depth estimation from oriental landscape painting images, we propose a novel framework that consists of two-step Image-to-Image translation method with CLIP-based image matching at the front end to predict the real scene image that best matches with the given oriental landscape painting image.
no code implementations • 9 Jan 2024 • Dongeon Kim, YeongHyeon Park
Recent studies try to use hyperspectral imaging (HSI) to detect foreign matters in products because it enables to visualize the invisible wavelengths including ultraviolet and infrared.
no code implementations • 6 Oct 2023 • YeongHyeon Park, Sungho Kang, Myung Jin Kim, Yeonho Lee, Hyeong Seok Kim, Juneho Yi
To enhance the UAD performance, reconstruction-by-inpainting based methods have recently been investigated, especially on the masking strategy of suspected defective regions.
Ranked #66 on Anomaly Detection on MVTec AD
Self-Supervised Anomaly Detection Supervised Anomaly Detection +1
no code implementations • 28 Aug 2023 • YeongHyeon Park, Sungho Kang, Myung Jin Kim, Hyeonho Jeong, Hyunkyu Park, Hyeong Seok Kim, Juneho Yi
In contrast, we note that containing of generalization ability in reconstruction can also be obtained simply from steep-shaped loss landscape.
no code implementations • 10 Jul 2023 • YeongHyeon Park, UJu Gim, Myung Jin Kim
However, the edge computers will have small data storage but we need to store the collected audio samples for a long time in order to update existing models or develop a novel method.
no code implementations • 20 Oct 2022 • YeongHyeon Park, Myung Jin Kim, Won Seok Park
Accurately extracting driving events is the way to maximize computational efficiency and anomaly detection performance in the tire frictional nose-based anomaly detection task.
no code implementations • 15 Jan 2022 • YeongHyeon Park
For the training anomaly detection model, the mean squared error (MSE) function is adopted widely.
no code implementations • 5 Jan 2022 • UJu Gim, YeongHyeon Park
We train normal and abnormal data as a feature that has a strong distinction among the features of imbalanced data.
no code implementations • 14 Dec 2021 • YeongHyeon Park, JoonSung Lee, Myung Jin Kim, Wonseok Park
Foreign substances on the road surface, such as rainwater or black ice, reduce the friction between the tire and the surface.
no code implementations • 22 Nov 2021 • YeongHyeon Park, JongHee Jung
As a result, the computational cost of the neural network is reduced up to 1 over 25 compared to the conventional models and the anomaly detection performance is improved by up to 7. 72%.
no code implementations • 7 Jul 2021 • JoonSung Lee, YeongHyeon Park
A model trained with normal data generates a larger restoration error for abnormal data.
no code implementations • 9 Apr 2021 • YeongHyeon Park, JoonSung Lee, Wonseok Park
Image classification technology and performance based on Deep Learning have already achieved high standards.
1 code implementation • 24 Mar 2021 • YeongHyeon Park, JongHee Jung
We conclude that NCAE as a cutting-edge model for road surface anomaly detection with 4. 20\% higher AUROC and 2. 99 times faster decision than before.
no code implementations • 18 Dec 2019 • YeongHyeon Park, Il Dong Yun, Si-Hyuck Kang
The Coronary Artery Occlusion (CAO) acutely comes to human, and it highly threats the human's life.
no code implementations • 2 Dec 2019 • YeongHyeon Park, Won Seok Park, Yeong Beom Kim
World Health Organization (WHO) provides the guideline for managing the Particulate Matter (PM) level because when the PM level is higher, it threats the human health.
1 code implementation • IEEE Access 2019 • YeongHyeon Park, Il Dong Yun, Si-Hyuck Kang
We mostly focus on enhancing the detecting performance using a preprocessing technique.
1 code implementation • MDPI 2018 • YeongHyeon Park, Il Dong Yun
Surface Mounted Device (SMD) assembly machine manufactures various products on a flexible manufacturing line.
no code implementations • 17 Jul 2018 • YeongHyeon Park, Il Dong Yun
Finally, we experimentally confirmed that the model performs better when the model restores the current sequence, rather than predict the future sequence.