Ho-Jin Choi

Prof. Ho-Jin Choi

School of Computing, Korea Advanced Institute of Science and Technology, Daejeon 34141, Korea

Research Focus & Metrics

Academic Career Length

15Years (2012-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
12 October 2022 in The Journal of Supercomputing
Verified
Journal Article
Open access
PDF file
MZMahe Zabin
Mahe Zabin
Ho-Jin ChoiHo-Jin Choi
Ho-Jin Choi
Jia UddinJia Uddin
Jia Uddin
Early-stage fault detection has become an indispensable part of modern industry to prevent potential hazards or sudden hindrances to the production process. With the advent of deep learning (DL) applications in several fields, DL models have been used to classify faults in specific environments. Uniform texture extraction has been performed using transformed-signal processing techniques and deep transfer learning (DTL) architectures in a few studies. Traditional signal processing techniques encounter difficulties in extracting distinct fault features due to the nonlinear and non-stationary nature of the time-series fault data. In this paper, a hybrid DTL architecture comprising a deep convolutional neural network and long short-term memory layers for extracting both temporal and spatial features enhanced by Hilbert transform 2D images is presented. Three standard audio sound fault datasets comprising the malfunctioning industrial machine investigation and inspection dataset, toy anomaly detection in machine operating sounds dataset, and machinery failure prevention technology bearing vibration fault dataset with various loads and noisy environments were utilized in the experimental evaluation. The proposed model with an input size of 32 × 32 achieved an average F1 score of 0.998 on the tested datasets. The implementation of transfer learning using the three benchmark datasets resulted in the highest accuracy of the proposed model and over fivefold reduction in the training epochs. In addition, the proposed model outperformed the state-of-art models in accuracy in various environments.

10 December 2018 in Peer-to-Peer Networking and Applications
Verified
Journal Article
Closed access
PDF file
ZLZhun Li
Zhun Li
BKByungSoo Ko
ByungSoo Ko
Ho-Jin ChoiHo-Jin Choi
Ho-Jin Choi

17 May 2021 in Sensors
Verified
Journal Article
Open access
PDF file
SISanghyuk Im
Sanghyuk Im
JHJonghwan Hyeon
Jonghwan Hyeon
EREun Young Rha
Eun Young Rha
JLJangHyeon Lee
JangHyeon Lee
Ho-Jin ChoiHo-Jin Choi
Ho-Jin Choi
Tae-Jung KimTae-Jung Kim
Tae-Jung Kim
Diffuse gliomas are the most common primary brain tumors and they vary considerably in their morphology, location, genetic alterations, and response to therapy. In 2016, the World Health Organization (WHO) provided new guidelines for making an integrated diagnosis that incorporates both morphologic and molecular features to diffuse gliomas. In this study, we demonstrate how deep learning approaches can be used for an automatic classification of glioma subtypes and grading using whole-slide images that were obtained from routine clinical practice. A deep transfer learning method using the ResNet50V2 model was trained to classify subtypes and grades of diffuse gliomas according to the WHO’s new 2016 classification. The balanced accuracy of the diffuse glioma subtype classification model with majority voting was 0.8727. These results highlight an emerging role of deep learning in the future practice of pathologic diagnosis.

26 September 2023 in Journal of Big Data
Verified
Journal Article
Open access
PDF file
MZMahe Zabin
Mahe Zabin
AKAnika Nahian Binte Kabir
Anika Nahian Binte Kabir
MKMuhammad Khubayeeb Kabir
Muhammad Khubayeeb Kabir
Ho-Jin ChoiHo-Jin Choi
Ho-Jin Choi
JUJia Uddin
Jia Uddin
Automated detection of defects on metal surfaces is crucial for ensuring quality control. However, the scarcity of labeled datasets for emerging target defects poses a significant obstacle. This study proposes a self-supervised representation-learning model that effectively addresses this limitation by leveraging both labeled and unlabeled data. The proposed model was developed based on a contrastive learning framework, supported by an augmentation pipeline and a lightweight convolutional encoder. The effectiveness of the proposed approach for representation learning was evaluated using an unlabeled pretraining dataset created from three benchmark datasets. Furthermore, the performance of the proposed model was validated using the NEU metal surface-defect dataset. The results revealed that the proposed method achieved a classification accuracy of 97.78%, even with fewer trainable parameters than the benchmark models. Overall, the proposed model effectively extracted meaningful representations from unlabeled image data and can be employed in downstream tasks for steel defect classification to improve quality control and reduce inspection costs.

1 June 2014 in Methods
Verified
Journal Article
Closed access
BJByeong-Soo Jeong
Byeong-Soo Jeong
ABA.T.M. Golam Bari
A.T.M. Golam Bari
MRMst. Rokeya Reaz
Mst. Rokeya Reaz
SJSeokhee Jeon
Seokhee Jeon
CLChae-Gyun Lim
Chae-Gyun Lim
Ho-Jin ChoiHo-Jin Choi
Ho-Jin Choi

1 March 2016 in IET Computer Vision
Verified
Journal Article
Open access
PDF file
NANur Aziza Azis
Nur Aziza Azis
YJYoung‐Seob Jeong
Young‐Seob Jeong
Ho-Jin ChoiHo-Jin Choi
Ho-Jin Choi
YIYoussef Iraqi
Youssef Iraqi
Existing studies in skeleton‐based action recognition mainly utilise skeletal data taken from a single camera. Since the quality of skeletal tracking of a single camera is noisy and unreliable, however, combining data from multiple cameras can improve the tracking quality and hence increase the recognition accuracy. In this study, the authors propose a method called weighted averaging fusion which merges skeletal data of two or more camera views. The method first evaluates the reliability of a set of corresponding joints based on their distances to the centroid, then computes the weighted average of selected joints, that is, each joint is weighted by the overall reliability of the camera reporting the joint. Such obtained, fused skeletal data are used as the input to the action recognition step. Experiments using various frame‐level features and testing schemes show that more than 10% improvement can be achieved in the action recognition accuracy using these fused skeletal data as compared with the single‐view case.