Minkyu Lee
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Minkyu Lee has a total of 19 co-authors. Here are the top co-authors by number of shared publications:
- Eungyeol Song
Research and Development Department, Codevision Inc., Seoul 03722, Republic of Korea
2 shared publications Chaebeen Kwon
Nanobio Device Laboratory School of Electrical and Electronic Engineering Yonsei University 50 Yonsei‐ro, Seodaemun‐gu Seoul 03722 Republic of Korea
1 shared publication- Chihyeong Won
Division of Engineering and Applied Science, California Institute of Technology, Pasadena, USA
1 shared publication - Dongsoo Han
School of Computing, Korea Advanced Institute of Science and Technology, Daejeon, South Korea
1 shared publication - Jaehong Lee
Department of Robotics and Mechatronics Engineering, DGIST, Daegu, Republic of Korea
1 shared publication - Ji-Soo Song
Department of Pediatric Dentistry, Dental Research Institute, School of Dentistry, Seoul National University, Seoul, Republic of Korea
1 shared publication
Latest Publications
- Open Access
- 27 June 2023
- Published by Wiley in Advanced Functional Materials
- Vol. 33 (26)
Multi‐Functional and Stretchable Thermoelectric Bi2Te3 Fabric for Strain, Pressure, and Temperature‐Sensing (Adv. Funct. Mater. 26/2023)
- 27 June 2023
- Published by Wiley in Advanced Functional Materials
- Vol. 33 (26)
Open AccessMulti‐Functional and Stretchable Thermoelectric Bi2Te3 Fabric for Strain, Pressure, and Temperature‐Sensing (Adv. Funct. Mater. 26/2023)
- Open Access
- 16 January 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 7, 11554-11564
Sampling Operator to Learn the Scalable Correlation Filter for Visual Tracking
The correlation filter is suitable for tracking on account of its low computational complexity and promising performance. However, the number of available training samples is limited to the filter size, and the lack of samples leads to poor generalization. Moreover, spectral leakage degrades the filter quality. In this paper, we therefore propose a sampling operator for learning a scalable correlation filter in an enlarged window, whose size is larger than the object size. The scalable filter encodes the sparse frequency representation to reconstruct a larger filter with zeros outside of the object in the spatial domain. The sampling operator, which is composed of windowing and sampling operations, enables learning the scalable filter from a large window, and it suppresses spectral leakage. Our method was evaluated on OTB- 100, TC-128, and UAV-123 datasets and achieved promising results in terms of precision and success rates.
- 16 January 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 7, 11554-11564
Open AccessSampling Operator to Learn the Scalable Correlation Filter for Visual Tracking
The correlation filter is suitable for tracking on account of its low computational complexity and promising performance. However, the number of available training samples is limited to the filter size, and the lack of samples leads to poor generalization. Moreover, spectral leakage degrades the filter quality. In this paper, we therefore propose a sampling operator for learning a scalable correlation filter in an enlarged window, whose size is larger than the object size. The scalable filter encodes the sparse frequency representation to reconstruct a larger filter with zeros outside of the object in the spatial domain. The sampling operator, which is composed of windowing and sampling operations, enables learning the scalable filter from a large window, and it suppresses spectral leakage. Our method was evaluated on OTB- 100, TC-128, and UAV-123 datasets and achieved promising results in terms of precision and success rates.
- Open Access
- 6 December 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 7, 284-292
CarvingNet: Content-Guided Seam Carving Using Deep Convolution Neural Network
We propose an improved content-aware image resizing method that uses deep learning. The proposed method is extended from seam carving, which is another image resizing method. Seam carving uses the energy map from an image. It also removes a seam where the energy is the minimum. We propose a method for creating a deep energy map using an encoder—decoder convolution neural network. A deep energy map preserves important parts or boundaries in an image, without distortion. Furthermore, it has the characteristic that uniform intensity of edges is displayed for all images. Four well-known resizing methods and our proposed method were evaluated in terms of aspect ratio similarity. In such an objective evaluation, the proposed method demonstrated better results, compared to the other four algorithms. Our proposed method can reduce the size of an image without damaging the overall structure or losing important information in the image.
- 6 December 2018
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 7, 284-292
Open AccessCarvingNet: Content-Guided Seam Carving Using Deep Convolution Neural Network
We propose an improved content-aware image resizing method that uses deep learning. The proposed method is extended from seam carving, which is another image resizing method. Seam carving uses the energy map from an image. It also removes a seam where the energy is the minimum. We propose a method for creating a deep energy map using an encoder—decoder convolution neural network. A deep energy map preserves important parts or boundaries in an image, without distortion. Furthermore, it has the characteristic that uniform intensity of edges is displayed for all images. Four well-known resizing methods and our proposed method were evaluated in terms of aspect ratio similarity. In such an objective evaluation, the proposed method demonstrated better results, compared to the other four algorithms. Our proposed method can reduce the size of an image without damaging the overall structure or losing important information in the image.
- Open Access
- 1 June 2007
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 58-63
A Probability-Based Prediction Framework for Stress Identification
In this paper, we propose a probability-based prediction framework for stress identification. The framework predicts the stress index of someone using vital signals and symptoms. The framework largely consists of learning stage and prediction stage. In learning stage, once vital signals and symptoms of stress group and normal group are available individually, a feature space is constructed. Vital signals and symptoms gathered from pre-identified subjects form the feature space. The framework constructs a matrix for each group which holding appearance frequencies of vital signals and symptom pair that is found in the feature space. Another matrix, SCAP matrix, is constructed using two matrices. It contains probabilities that indicate a certain vital signals-symptom pair may be found in stress group and the framework prepares weight vectors in order to direct the use of vital signals and symptoms in personalized manner. In prediction stage, a probability equation is devised based on SCAP matrix and the weight vectors. Once the vital signal and symptom pair of someone is available, the probability that indicates a person may belong to a stress group is computed using the probability equation. In this paper, we conduct a simple validation test for this framework using vital signal and symptom data obtained from 29 subjects. The method is revealed to have high potential to improve probability of stress identification further after more learning data is accumulated.
- 1 June 2007
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 58-63
Open AccessA Probability-Based Prediction Framework for Stress Identification
In this paper, we propose a probability-based prediction framework for stress identification. The framework predicts the stress index of someone using vital signals and symptoms. The framework largely consists of learning stage and prediction stage. In learning stage, once vital signals and symptoms of stress group and normal group are available individually, a feature space is constructed. Vital signals and symptoms gathered from pre-identified subjects form the feature space. The framework constructs a matrix for each group which holding appearance frequencies of vital signals and symptom pair that is found in the feature space. Another matrix, SCAP matrix, is constructed using two matrices. It contains probabilities that indicate a certain vital signals-symptom pair may be found in stress group and the framework prepares weight vectors in order to direct the use of vital signals and symptoms in personalized manner. In prediction stage, a probability equation is devised based on SCAP matrix and the weight vectors. Once the vital signal and symptom pair of someone is available, the probability that indicates a person may belong to a stress group is computed using the probability equation. In this paper, we conduct a simple validation test for this framework using vital signal and symptom data obtained from 29 subjects. The method is revealed to have high potential to improve probability of stress identification further after more learning data is accumulated.