23 citations
- Korea Advanced Institute of Science and TechnologyKR2 papers
- Pohang University of Science and TechnologyKR2 papers
- Busitema UniversityUG1 paper
- Hyundai Mobis (South Korea)KR1 paper
- Korea Research Institute of Standards and ScienceKR1 paper
- Korea University of Science and TechnologyKR1 paper
- Mbarara University of Science and TechnologyUG1 paper
- Norwegian University of Science and TechnologyNO1 paper
- Pohang TechnoPark (South Korea)KR1 paper
- Tongji UniversityCN1 paper
- University of Uyo Teaching HospitalNG1 paper
- Virginia TechUS1 paper
24 papers
Holi-DETR: Holistic Fashion Item Detection Leveraging Contextual Information
Youngchae Kwon, Jinyoung Choi, Injung Kim
Fashion item detection is challenging due to the ambiguities introduced by the highly diverse appearances of fashion items and the similarities among item subcategories. To address…
Item Region-based Style Classification Network (IRSN): A Fashion Style Classifier Based on Domain Knowledge of Fashion Experts
Jinyoung Choi, Youngchae Kwon, Injung Kim
Fashion style classification is a challenging task because of the large visual variation within the same style and the existence of visually similar styles. Styles are expressed no…
MF-LPR: Multi-Frame License Plate Image Restoration and Recognition using Optical Flow
Kihyun Na, Junseok Oh, Youngkwan Cho +4
License plate recognition (LPR) is important for traffic law enforcement, crime investigation, and surveillance. However, license plate areas in dash cam images often suffer from l…
Data Augmentation With Back translation for Low Resource languages: A case of English and Luganda
Richard Kimera, Dongnyeong Heo, Daniela N. Rim +1
In this paper,we explore the application of Back translation (BT) as a semi-supervised technique to enhance Neural Machine Translation(NMT) models for the English-Luganda language…
PECI-Net: Bolus segmentation from video fluoroscopic swallowing study images using preprocessing ensemble and cascaded inference
Dougho Park, Younghun Kim, Harim Kang +7
Bolus segmentation is crucial for the automated detection of swallowing disorders in videofluoroscopic swallowing studies (VFSS). However, it is difficult for the model to accurate…
Enhanced Labeling Technique for Reddit Text and Fine-Tuned Longformer Models for Classifying Depression Severity in English and Luganda
Richard Kimera, Daniela N. Rim, Joseph Kirabira +2
Depression is a global burden and one of the most challenging mental health conditions to control. Experts can detect its severity early using the Beck Depression Inventory (BDI) q…