• Journal of the HCI Society of Korea
• /
• v.14 no.2
• /
• pp.49-60
• /
• 2019

With the appearance of self-driving cars and electric cars, the automobile industry is rapidly changing. In the midst of these changes, HMI studies are becoming more important as to how the driver obtains safety and convenience with controlling the vehicle. This study sought to understand how automobile manufacturers understand the driving situation, and how they define and limit driver interaction. For this, prior studies about HMI were reviewed and 15 participants performed an on-road study to drive vehicles from five manufacturers with using their interfaces. The results of the study confirmed that buttons and switches that are easily controlled by the user while driving were different from manufacturer to manufacturer. And there are some buttons that are more intensively controlled and others that are difficult to control while driving. It was able to derive 'selection and concentration' from Audi's vehicle, 'optimization of the driving ' from BMW's, 'simple and minimize' from Benz's vehicle, 'remove the manual distraction' from the vehicle of Lexus, and 'visual stability' from KIA's vehicle as the distinctive keywords for the HMI. This shows that each manufacturer has a different definition and interpretation of the driver's driving control area. This study has a distinct value in that it has identified the characteristics of vehicle-specific HMI in actual driving conditions, which is not apparent in appearance. It is expected that this research approach can be useful to see differences in interaction through actual driving despite changes in driving environment such as vehicle platooning and self-driving technology.

• The Journal of Society for e-Business Studies
• /
• v.24 no.3
• /
• pp.35-51
• /
• 2019

The two main goals of this study are to derive independent factors affecting the success rate of crowdfunding and to empirically analyze the variation of independent factors' effects on the success of crowdfunding by industry (Internet, culture/art, manufacturing/distribution), and funding type (stock type, bond type). To identify the success factors of crowdfunding for invigoration and strategic utilization, first, several variables were refined after interviews with experts and platform operators with investment experiences in numerous crowdfunding projects. Then, independent factors affecting project involvement were categorized as follows: a characteristic of project, participant activity, and enterprise. Also, the results derived from the influence of independent variables on crowdfunding after moderating effects were driven. Selected independent factors in this study are as follows: crowdfunding period, target amount, visual contents, minimum account money, number of comments, number of SNS followers, level of interest, financial Statement disclosure, investment attraction, venture company, intellectual property rights disclosure, and business operation period. Selected moderating factors in this study are as follows: industry (Internet, culture/art, manufacturing/distribution), and funding type (stock type, bond type). In conclusion, a discussion of the academical and practical implications and a suggestion of directions for further research are explained.

• Phonetics and Speech Sciences
• /
• v.11 no.1
• /
• pp.41-49
• /
• 2019

An end-to-end speech recognition model consisting of a single integrated neural network model was recently proposed. The end-to-end model does not need several training steps, and its structure is easy to understand. However, it is difficult to understand how the model recognizes speech internally. In this paper, we visualized and analyzed the attention-based end-to-end model to elucidate its internal mechanisms. We compared the acoustic model of the BLSTM-HMM hybrid model with the encoder of the end-to-end model, and visualized them using t-SNE to examine the difference between neural network layers. As a result, we were able to delineate the difference between the acoustic model and the end-to-end model encoder. Additionally, we analyzed the decoder of the end-to-end model from a language model perspective. Finally, we found that improving end-to-end model decoder is necessary to yield higher performance.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.20 no.7
• /
• pp.607-612
• /
• 2019

The softness of the crashpad part is one of the important factors which affect the interior perceived quality of the vehicle interior. And while improving the softness of the crashpad part, every effort to lower the production cost has been going on. The PU foaming process for the crashpad part depends on the understanding of a lot of processes, tools and material properties. Therefore, to achieve the requirement of the customer for the interior part's visual quality, the integrated design techniques are investigated to correlate the processes, tool design, material design and the computer aided analysis. In this paper, IMG (In Mold Grain) designed concept is firstly developed to integrate the skin preforming, plastic injection molding of the substrate and the foaming process in a tool within reduced processes. Through the application of this technology, softness of crashpad is improved by 40% compared to the conventional vacuum molding method, and the existing process is reduced by 50% by integrating the injection process and the manufacturing process. And by integrating the injection mold and the skin mold and removing the foaming mold, the number of molds are reduced from 3 to 1, resulting in 20% reduction in the cost of applying a medium-sized passenger car.

• Science of Emotion and Sensibility
• /
• v.24 no.3
• /
• pp.41-60
• /
• 2021

The purpose of this research was to determine whether multimodal cues in an online shopping environment could enhance tactile consumer mental imagery, purchase intentions, and attitudes towards an apparel product. One limitation of online retail is that consumers are unable to physically touch the items. However, as tactile information plays an important role in consumer decisions especially for apparel products, this study investigated the effects of multimodal cues on overcoming the lack of tactile stimuli. In experiment 1, to explore the product, the participants were randomly assigned to four conditions; picture only, video without sound, video with corresponding sound, and video with discordant sound; after which tactile mental imagery vividness, ease of imagination, attitude, and purchase intentions were measured. It was found that the video with discordant sound had the lowest average scores of all dependent variables. A within-participants design was used in experiment 2, in which all participants explored the same product in the four conditions in a random order. They were told that they were visiting four different brands on a price comparison web site. After the same variables as in experiment 1, including the need for touch, were measured, the repeated measures ANCOVA results revealed that compared to the other conditions, the video with the corresponding sound significantly enhanced tactile mental imagery vividness, attitude, and purchase intentions. However, the discordant condition had significantly lower attitudes and purchase intentions. The dual mediation analysis also revealed that the multimodal cue conditions significantly predicted attitudes and purchase intentions by sequentially mediating the imagery vividness and ease of imagination. In sum, vivid tactile mental imagery triggered using audio-visual stimuli could have a positive effect on consumer decision making by making it easier to imagine a situation where consumers could touch and use the product.

• Journal of the Korea Society of Computer and Information
• /
• v.26 no.9
• /
• pp.155-166
• /
• 2021

Recently, digital health care technology is spreading and developing in various fields. Therefore, in this paper, we realized that the field to which digital healthcare technology is not applied is the field of optometry, and implemented a digital healthcare optometry system for precise lens manufacturing. A device called Phoroptor is used to manufacture the lens, and this device sets the lens by measuring the visual acuity of the person who requested the glasses. And when the person to be measured wears glasses, a device called a PD meter is used to align the pupil center and lens focus. However, there is a limit to the convenience of precise lens production and optometry due to the absence of a database and program that can accumulate and analyze the PD measurement error, inconvenience and error due to manual control of the Phoroptor, and optometric information. Therefore, in this paper, PD meter design for more accurate PD measurement, Phoroptor design and Phoroptor control application design for automatic Phoroptor control, and a database and analysis program that automatically set lenses using optometry information for each subject had been designed. Based on this, ultimately, a digital healthcare optometry system using an optometry database has been implemented.

• KIPS Transactions on Software and Data Engineering
• /
• v.9 no.12
• /
• pp.403-410
• /
• 2020

In addition to the research on noise removal and super-resolution using the data restoration (Output result) function of Autoencoder, research on the performance improvement of clustering using the dimension reduction function of autoencoder are actively being conducted. The clustering function and data restoration function using Autoencoder have common points that both improve performance through the same learning. Based on these characteristics, this study conducted an experiment to see if the autoencoder model designed to have excellent data recovery performance is superior in clustering performance. Skip connection technique was used to design autoencoder with excellent data recovery performance. The output result performance and clustering performance of both autoencoder model with Skip connection and model without Skip connection were shown as graph and visual extract. The output result performance was increased, but the clustering performance was decreased. This result indicates that the neural network models such as autoencoders are not sure that each layer has learned the characteristics of the data well if the output result is good. Lastly, the performance degradation of clustering was compensated by using both latent code and skip connection. This study is a prior study to solve the Hanja Unicode problem by clustering.

• Korean Journal of Remote Sensing
• /
• v.37 no.3
• /
• pp.577-589
• /
• 2021

Drone mapping systems can be applied to many fields such as disaster damage investigation, environmental monitoring, and construction process monitoring. To integrate individual sensors attached to a drone, it was essential to undergo complicated procedures including time synchronization. Recently, a variety of composite sensors are released which consist of visual sensors and GPS/INS. Composite sensors integrate multi-sensory data internally, and they provide geotagged image files to users. Therefore, to use composite sensors in drone mapping systems, mapping accuracies from composite sensors should be examined. In this study, we analyzed the mapping accuracies of a composite sensor, focusing on the data acquisition area and pre-calibration effect. In the first experiment, we analyzed how mapping accuracy varies with the number of ground control points. When 2 GCPs were used for mapping, the total RMSE has been reduced by 40 cm from more than 1 m to about 60 cm. In the second experiment, we assessed mapping accuracies based on whether pre-calibration is conducted or not. Using a few ground control points showed the pre-calibration does not affect mapping accuracies. The formation of weak geometry of the image sequences has resulted that pre-calibration can be essential to decrease possible mapping errors. In the absence of ground control points, pre-calibration also can improve mapping errors. Based on this study, we expect future drone mapping systems using composite sensors will contribute to streamlining a survey and calibration process depending on the data acquisition circumstances.

• Clinics in Shoulder and Elbow
• /
• v.24 no.1
• /
• pp.21-26
• /
• 2021

Background: To evaluate the demographics, clinical and radiographic features of calcific tendinitis of the shoulder in the Korean population, specifically focusing on the incidence of coexisting rotator cuff tear. Methods: Between October 2014 and January 2015, we performed a prospective multicenter study with 506 patients from 11 training hospitals in Korea. We collected data of demographics and radiographic analysis based on simple radiographs, clinical assessments based on visual analog scale (VAS) and the American Shoulder Elbow Surgeons (ASES) score, and treatment modalities that are used currently. We also evaluated coexisting rotator cuff tear by ultrasonography (US) or magnetic resonance imaging (MRI) images. Results: There were 402 female patients (79%) with mean age of 55 years (range, 31-87 years). Mean duration of symptoms was 16 months. Mean size of calcific materials was 11.4 mm (range, 0-35 mm). Mean value of VAS and ASES scores were 6.5 (range, 1-10) and 47 (range, 8-95), respectively. Of 383 patients (76%), 59 (15%) had rotator cuff tear including 15 full-thickness tears on US or MRI. Patients with rotator cuff tears were significantly associated with older age, recurrent symptoms, menstrual disorders in females, and having undergone calcification removal surgery and rotator cuff repair (all p<0.05). Conclusions: This study reported demographic, radiographic, and clinical features of calcific tendinitis of the shoulder in Korean population, which were not different from those of Western population. Coexisting rotator cuff tear was found with 15% incidence in this large series, suggesting that further radiographic study to evaluate rotator cuff tear might be needed in some calcific tendinitis patients of older age and presenting with recurrent symptoms.

• Journal of the Korean Geosynthetics Society
• /
• v.19 no.4
• /
• pp.41-50
• /
• 2020

Since cracks in concrete structures expedite corrosion of reinforced concrete over a long period of time, regular on-site inspections are essential to ensure structural usability and prevent degradation. Most of the safety inspections of facilities rely on visual inspection with naked eye, so cost and time consuming are severe, and the reliability of results differs depending on the inspector. In this study, a portable measuring device that can be used for safety diagnosis and maintenance was developed as a device that measures the width and length of concrete cracks through image analysis of cracks photographed with a camera. This device captures the cracks found within a close distance (3 m), and accurately calculates the unit pixel size by laser distance measurement, and automatically calculates the crack length and width with the image processing algorithm developed in this study. In measurement results using the crack image applied to the experiment, the measurement of the length of a 0.3 mm crack within a distance of 3 m was possible with a range of about 10% error. The crack width showed a tendency to be overestimated by detecting surrounding pixels due to vibration and blurring effect during the binarization process, but it could be effectively corrected by applying the crack width reduction function.