A5101447885- Image Processing Techniques and Applications
- Industrial Vision Systems and Defect Detection
- Intellectual Property and Patents
- Integrated Circuits and Semiconductor Failure Analysis
- Advanced Image Processing Techniques
- Innovation Diffusion and Forecasting
- Image and Signal Denoising Methods
- Complex Network Analysis Techniques
- Innovation and Knowledge Management
- Business Strategy and Innovation
- Firm Innovation and Growth
- CCD and CMOS Imaging Sensors
- Innovation Policy and R&D
- Technology Assessment and Management
- Domain Adaptation and Few-Shot Learning
- Bone Tissue Engineering Materials
- Computational Drug Discovery Methods
- Welding Techniques and Residual Stresses
- Advanced Vision and Imaging
- Face and Expression Recognition
- Big Data and Business Intelligence
- Microstructure and Mechanical Properties of Steels
- Advanced Neural Network Applications
- Image Enhancement Techniques
- Image and Object Detection Techniques
Hanyang University
2019-2025
Seoul National University
2023
Rutgers, The State University of New Jersey
2013-2020
Korea Advanced Institute of Science and Technology
2020
Korea Institute of Science & Technology Information
2015-2017
Doosan Heavy Industries & Construction (South Korea)
2013
Global Core Research Center for Ships and Offshore Plants
2012
Pusan National University
2012
Samsung (South Korea)
2010
This paper reviews the NTIRE 2020 challenge on perceptual extreme super-resolution with focus proposed solutions and results.The task was to super-resolve an input image a magnification factor ×16 based set of prior examples low corresponding high resolution images.The goal is obtain network design capable produce results best quality similar ground truth.The track had 280 registered participants, 19 teams submitted final results.They gauge state-of-the-art in single superresolution....
Adopting smart technologies for supply chain management leads to higher profits. The manufacturer and retailer are two players, where the is unreliable may not send accurate demand information manufacturer. As an advanced technology, Radio Frequency Identification (RFID) implemented track trace each product’s movement on a real-time basis in inventory. It takes this management. This research proposes Machine Learning (ML) approach on-demand forecasting under Using Long-Short-Term Memory...
Recent advances in large language models (LLMs) have accelerated the development of conversational agents capable generating human-like responses. Since psychiatric assessments typically involve complex interactions between psychiatrists and patients, there is growing interest developing LLM-based assessment (PACAs) that aim to simulate role clinical evaluations. However, standardized methods for benchmarking appropriateness PACAs' interaction with patients still remain underexplored. Here,...
Effectively ranking patents in outlierness a patent citation network is crucial task for analysis, including as it relates to technological opportunity discovery (TOD). Previous studies the area of TOD focus on textual data. In this paper, we introduce new approach that addresses via outlierness, leveraging both attributes and citations. We propose following characteristics outliers: 1) not highly clustered with other patents; 2) low node centrality within network; 3) similarity network....
In semiconductor manufacturing processes, monitoring the quality of wafers is one most important steps to quickly detect process faults and significantly reduce yield loss. Fail bit maps (FBMs) represent failed cell count during wafer functional tests have been popularly used as diagnosis tools in for monitoring, root cause analysis, improvements. However, visual inspection costly time-consuming. Therefore, this paper proposes an automated classification procedure failure patterns on FBMs...
Classification of defect chip patterns is one the most important tasks in semiconductor manufacturing process. During final stage process just before release, engineers must manually classify and summarise information chips from a number wafers that can aid diagnosing root causes failures. Traditionally, several learning algorithms have been developed to on wafer maps. However, them focused single bin map based certain features. The objective this study propose novel approach multiple maps...
Defects on semiconductor wafers are not uniformly distributed, but tend to cluster. These spatial defect patterns contain useful information about issues during integrated circuit fabrication. Promptly detecting abnormal is an important way increase yield and product quality. However, research identifying has focused only flash memory with a single wafer map. No procedure available for dynamic random access (DRAM) multiple maps. This paper proposes new step-down randomness test abnormalities...
Identifying competitors is an essential first step in creating a competitive advantage to generate superior business performance. Diverse methods for competitor identification have been introduced the literature, but financial transactions are overlooked these models even though such among firms, as basic economic unit of interfirm activities, contain critical information. This article proposes novel method firm-level using network constructed from firms. To identify by intensity...
This paper reviews the second AIM learned ISP challenge and provides description of proposed solutions results. The participating teams were solving a real-world RAW-to-RGB mapping problem, where to goal was map original low-quality RAW images captured by Huawei P20 device same photos obtained with Canon 5D DSLR camera. considered task embraced number complex computer vision subtasks, such as image demosaicing, denoising, white balancing, color contrast correction, demoireing, etc. target...
This study presents a novel approach for identifying new business opportunities by analyzing the linkage between patents and trademarks leveraging text analytics. Initially, we utilize topic modeling to analyze descriptions of goods services in trademarks, with particular focus on that do not share similar group codes. Using Latent Dirichlet Allocation (LDA) model, are segmented into multiple groups based similarities. Subsequently, define areas measuring their similarity industry...
The selection of group features is a critical aspect in reducing model complexity by choosing the most essential features, while eliminating less significant ones. existing feature methods select set important without providing relative importance all features. Moreover, few consider process. This study introduces permutation-based approach specifically designed for high-dimensional multiclass datasets. Initially, least absolute shrinkage and operator (lasso) method was applied to eliminate...
This paper introduces the real image Super-Resolution (SR) challenge that was part of Advances in Image Manipulation (AIM) workshop, held conjunction with ECCV 2020. involves three tracks to super-resolve an input for $\times$2, $\times$3 and $\times$4 scaling factors, respectively. The goal is attract more attention realistic degradation SR task, which much complicated challenging, contributes real-world super-resolution applications. 452 participants were registered total, 24 teams...
Group feature selection methods select the important group features by removing irrelevant for reducing complexity of model. There are a few that used ranking techniques. Ranking provide relative importance each group. For this purpose, we developed sparse method based on dimension reduction technique high dimensional data. Firstly, applied relief to remove features. Secondly, extract new represents To end, reduce into single applying Fisher linear discriminant analysis (FDA) At last,...
A new type of artificial neural network to predict the reaction rate constant from two molecular structures. An explainable model was constructed using multi-head intermolecular attention technique.
More than 70% of the accidents that occur on offshore installations stem from hydrocarbon explosions and fires, which, because they involve blast effects heat, are extremely hazardous have serious consequences in terms human health, structural safety, surrounding environment. Blast barriers integral structures a typical topside module to protect personnel safety critical equipment by preventing escalation events caused explosions. Many researchers shown adequacy simple design tool commonly...
This study deals with the preventive maintenance optimization problem based on a reliability threshold. The conditional threshold is used instead of system Then, difference between two thresholds discussed. hybrid failure rate model employed to represent effect imperfect activities. Two strategies are proposed under types constraints. These constraints set consider cost-effective strategy and evaluate balancing point expected total cost reliability. objective determine optimal together...