• Proceedings of the Korea Information Processing Society Conference
• /
• 2021.11a
• /
• pp.1001-1004
• /
• 2021

Pain is an unpleasant experience for the patient. The recognition and assessment of pain help tailor the treatment to the patient, and they are also challenging in the medical. In this paper, we propose an approach for pain recognition through a deep neural network applied to pre-processed physiological. The proposed approach applies the idea of shortcut connections to concatenate the spatial information of a convolutional neural network and the temporal information of a recurrent neural network. In addition, our proposed approach applies the attention mechanism and achieves competitive performance on the BioVid Heat Pain dataset.

• East Asian mathematical journal
• /
• v.40 no.2
• /
• pp.267-286
• /
• 2024

This study investigates the perceptions toward prospective elementary teachers regarding the revised 2015 elementary mathematics curriculum. The aim is to understand the importance and implementation of the revised curriculum and provide implications for curriculum improvement in elementary teacher education institutions, using Interpretative Phenomenological Analysis (IPA). The research findings are as follows: Firstly, prospective elementary teachers perceived that the areas of the revised 2015 elementary mathematics curriculum that require particular focus are number and operations and data and probability. Secondly, they identified the specific elements within these areas that demand dedicated attention as follows: numbers up to four digits in number and operations, mixed calculations with natural numbers, shapes of solid figures, spatial sense of solid figures, comparison of quantities in measurement, etc. These findings can inform the improvement of the curriculum in elementary teacher education institutions.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.51 no.4
• /
• pp.110-122
• /
• 2014

As multimedia technologies develop, three-dimensional (3D) technologies are attracting increasing attention from researchers. In particular, video quality assessment (VQA) has become a critical issue in stereoscopic image/video processing applications. Furthermore, a human visual system (HVS) could play an important role in the measurement of stereoscopic video quality, yet existing VQA methods have done little to develop a HVS for stereoscopic video. We seek to amend this by proposing a 3D visual attention (3DVA) model which simulates the HVS for stereoscopic video by combining multiple perceptual stimuli such as depth, motion, color, intensity, and orientation contrast. We utilize this 3DVA model for pooling on significant regions of very poor video quality, and we propose no-reference (NR) stereoscopic VQA (SVQA) method. We validated the proposed SVQA method using subjective test scores from our results and those reported by others. Our approach yields high correlation with the measured mean opinion score (MOS) as well as consistent performance in asymmetric coding conditions. Additionally, the 3DVA model is used to extract information for the region-of-interest (ROI). Subjective evaluations of the extracted ROI indicate that the 3DVA-based ROI extraction outperforms the other compared extraction methods using spatial or/and temporal terms.

• Journal of Advanced Navigation Technology
• /
• v.27 no.2
• /
• pp.190-196
• /
• 2023

Recently, following the development of LIDAR technology which can detect distance from the object, the interest for LIDAR based 3D object detection network is getting higher. Previous networks generate inaccurate localization results due to spatial information loss during voxelization and downsampling. In this study, we propose an attention-based convergence method and a camera-LIDAR convergence system to acquire high-level features and high positional accuracy. First, by introducing the attention method into the Voxel-RCNN structure, which is a grid-based 3D object detection network, the multi-scale sparse 3D convolution feature is effectively fused to improve the performance of 3D object detection. Additionally, we propose the late-fusion mechanism for fusing outcomes in 3D object detection network and 2D object detection network to delete false positive. Comparative experiments with existing algorithms are performed using the KITTI data set, which is widely used in the field of autonomous driving. The proposed method showed performance improvement in both 2D object detection on BEV and 3D object detection. In particular, the precision was improved by about 0.54% for the car moderate class compared to Voxel-RCNN.

• Journal of Broadcast Engineering
• /
• v.27 no.3
• /
• pp.391-401
• /
• 2022

As deep learning-based object detection and recognition research have been developed recently, the scope of application to industry and real life is expanding. But deep learning-based systems in the construction system are still much less studied. Calculating materials in the construction system is still manual, so it is a reality that transactions of wrong volumn calculation are generated due to a lot of time required and difficulty in accurate accumulation. A fast and accurate automatic drawing recognition system is required to solve this problem. Therefore, we propose an AI-based automatic drawing recognition accumulation system that detects and recognizes steel materials in construction drawings. To accurately detect steel materials in construction drawings, we propose data augmentation techniques and spatial attention modules for improving small object detection performance based on YOLOv4. The detected steel material area is recognized by text, and the number of steel materials is integrated based on the predicted characters. Experimental results show that the proposed method increases the accuracy and precision by 1.8% and 16%, respectively, compared with the conventional YOLOv4. As for the proposed method, Precision performance was 0.938. The recall was 1. Average Precision AP_0.5 was 99.4% and AP_0.5:0.95 was 67%. Accuracy for character recognition obtained 99.9.% by configuring and learning a suitable dataset that contains fonts used in construction drawings compared to the 75.6% using the existing dataset. The average time required per image was 0.013 seconds in the detection, 0.65 seconds in character recognition, and 0.16 seconds in the accumulation, resulting in 0.84 seconds.

• Journal of Digital Convergence
• /
• v.20 no.5
• /
• pp.281-289
• /
• 2022

As the aging of our society gradually deepens, the quality of the elderly care environment based on the elderly care model of the Silver Town space is gradually becoming the focus of everyone's attention. This study mainly studies and discusses the visual and behavioral characteristics of the elderly. In order to pay more attention to the special needs of the elderly in order to optimize the spatial environment, the design of visual experience design in the living environment of the elderly in Silver Town is designed. explore in depth Through this, the environment of the space is optimized so that the elderly can enjoy their old age in a comfortable environment. This study is based on visual psychology, environmental psychology, gerontology and other theories. Through the collection of related literature and field research on the elderly, the function and overall combination of the living environment of the elderly in Silver Town is studied, and the environment is organized. Based on the behavioral and visual needs of middle-aged and older people, a design method was proposed to strengthen the visual connection in space. In terms of visual experience, the lighting, colors, and materials of the environment are studied. Through a combination of theory and research and experiments, it is concluded that the elderly prefer plants with warm colors, high brightness colors, and geometric patterns. The design principle and design method of the visual experience in the Silver Town space are summarized.

• Korean Institute of Interior Design Journal
• /
• v.26 no.2
• /
• pp.22-32
• /
• 2017

This research is to analyze the spatial information search features which shown by Eye fixation and movement and conducted eye tracking experiment for targeting sports shop spatial images which it are same but looks different. This is able to find out the eye movement feature according to placement of goods from the eye movement and movement distance of spatial visitor, and the result can be defined as following. First, the whole original-reverse left / right images have a higher number of observations in the [IN] area than in the [OUT] area. This is because after eye taking high observations in LA area of [IN] have been jump-over [OUT], performed search activities in low eye fixation without high eye fixation. Second, there was a difference in the frequency of the observation data as the composition of the images changed. The original image has been often fixed the eyes in LA area, and the one that has been observed for a long time is reverse left / right image. Also, fixation point was shown higher at the reverse left / right image as jump-over from [OUT] area to [IN] area. If LA area seen as reverse left / right image, it is located in right-hand side. The case where the dominant area is on the right side has a characteristic that the eye fixation is longer. This can be understand that the arrangement of products for attract the customer's attention in the commercial space might be more effective when it is on the right side. Third, the moving distance(IN ${\rightarrow}$ OUT) of the sight pointed to external from LA area was long in the both original-reverse left / right images, but it is no relation with search direction([IN${\rightarrow}$OUT] [IN${\rightarrow}$OUT]) of the sight. In other words, the sight that entered in LA area can be seen as visual perception activity for re-searching after big jump-over, in the case go in to outward (OUT area) after searching for more than certain time. The fact that the moving distance of eye is relatively short in the [IN ${\rightarrow}$ OUT] process considered as that the gaze that stays outside the LA area naturally enters in to LA area.

• The Journal of the Korea Contents Association
• /
• v.12 no.4
• /
• pp.186-197
• /
• 2012

The purpose of the thesis is to answer the question on how the visual image content, being the original content, should be adapted to and represented as a spatial content. The thesis focuses on adaptation as the key in the conversion process of visual image content to a themed space. There are many published studies dealing with storytelling, adaptation from books to movies and TV shows, or from movies to games and vice versa. On the contrary, when it comes to adaptation from visual image content to spatial content, noticeably few studies were done on the method, and fewer studies view adaptation as the prior step of storytelling. This study first defines adaptation, and then applies the methods of Gianetty and Dudley which is further incorporated into the conversion of visual image content into a themed space. It then turns the attention to the characteristics of themed spaces. A case study highlights that a themed space is a spatial representation of the story, image and action in the visual image content, and analyze the type of adaptation made. The study results draws two conclusions; adaptation must be carried out prior to the storytelling of the spatial content; and opposed to a third-person view of the visual image content, the main factor in a themed space is first-hand experience. Thus, the thesis suggests that conversion from visual image content to themed spaces are not merely imitative but is a full range of recreation of a new content. It is expected that more detailed analyses on the particulars will lead to feasible outcome on implementing various methods of adaptation and bring about effective conversions between the visual image contents and themed spaces.

• Progress in Medical Physics
• /
• v.27 no.3
• /
• pp.146-155
• /
• 2016

As the probability of exposure to radiation increases due to an increase in the use of radioisotopes and radiation generators, the importance of a radiation safety management field is being highlighted. We intend to help radiation workers with exposure management by identifying the degree of radiation exposure and contamination to determine an efficient method of radiation safety management. The personal exposure doses of the radiation workers at the Korea Institute of Radiological & Medical Sciences measured every quarter during a five-year period from Jan. 1, 2011 till Dec. 31, 2015 were analyzed using a TLD (thermoluminescence dosimeter). The spatial dose rates of radiation-controlled areas were measured using a portable radioscope, and the level of surface contamination was measured at weekly intervals using a piece of smear paper and a low background alpha/beta counter. Though the averages of the depth doses and the surface doses in 2012 increased from those in 2011 by about 14%, the averages were shown to have decreased every year after that. The exposure dose of 27 mSv in 2012 increased from that in 2011 in radiopharmaceutical laboratories and, in the case of the spatial dose rate, the rate of decrease in 2012 was shown to be similar to the annual trend of the whole institute. In the case of the surface contamination level, as the remaining radiation-controlled area with the exception of the I-131 treatment ward showed a low value less than $1.0kBq/m^2$, the annual trend of the I-131 treatment ward was shown to be similar to that of the entire institute. In conclusion, continuous attention should be paid to dose monitoring of the radiation-controlled areas where unsealed sources are handled and the workers therein.

• Journal of Cadastre & Land InformatiX
• /
• v.49 no.2
• /
• pp.57-66
• /
• 2019

Recently, as a part of the production of spatial information by smart cities, three-dimensional reproduction of structures for reverse engineering has been attracting attention. In particular, terrestrial LiDAR is mainly used for 3D reproduction of structures, and 3D reproduction research by UAS has been actively conducted. However, both technologies produce blind spots due to the shooting angle. This study deals with vertical structures. 3D model implemented through SfM-based image analysis technology using UAS and reproducibility and effectiveness of 3D models by terrestrial LiDAR-based laser scanning are examined. In addition, two 3D models are merged and reviewed to complement the blind spot. For this purpose, UAS based image is acquired for artificial rock wall, VCP and check point are set through GNSS equipment and total station, and 3D model of structure is reproduced by using SfM based image analysis technology. In addition, Through 3D LiDAR scanning, the 3D point cloud of the structure was acquired, and the accuracy of reproduction and completeness of the 3D model based on the checkpoint were compared and reviewed with the UAS-based image analysis results. In particular, accuracy and realistic reproducibility were verified through a combination of point cloud constructed from UAS and terrestrial LiDAR. The results show that UAS - based image analysis is superior in accuracy and 3D model completeness and It is confirmed that accuracy improves with the combination of two methods. As a result of this study, it is expected that UAS and terrestrial LiDAR laser scanning combination can complement and reproduce precise three-dimensional model of vertical structure, so it can be effectively used for spatial information construction, safety diagnosis and maintenance management.