• Phonetics and Speech Sciences
• /
• v.14 no.3
• /
• pp.1-9
• /
• 2022

The current study examined English-speaking adult learners' production and perception of L2 Korean stops (/t/ or /t'/ or /t^h/) to investigate whether the two modalities are linked in utilizing voice onset time (VOT) and fundamental frequency (F0) for the L2 sound distinction and how the learners' L2 proficiency mediates the relationship. Twenty-two English-speaking learners of Korean living in Seoul participated in the word-reading task of producing stop-initial words and the identification task of labelling CV stimuli synthesized to vary VOT and F0. Using logistic mixed-effects regression models, we quantified group- and individual-level weights of the VOT and F0 cues in differentiating the tense-lax, lax-aspirated, and tense-aspirated stops in Korean. The results showed that the learners as a group relied on VOT more than F0 both in production and perception (except the tense-lax pair), reflecting the dominant role of VOT in their L1 stop distinction. Individual-level analyses further revealed that the learners' L2 proficiency was related to their use of F0 in L2 production and their use of VOT in L2 perception. With this effect of L2 proficiency controlled in the partial correlation tests, we found a significant correlation between production and perception in using VOT and F0 for the lax-aspirated stop contrast. However, the same correlation was absent for the other stop pairs. We discuss a contrast-specific role of acoustic cues to address the non-uniform patterns of the production-perception link in the L2 sound learning context.

• KIPS Transactions on Software and Data Engineering
• /
• v.11 no.8
• /
• pp.325-330
• /
• 2022

This paper conducts the collection and experiment of datasets for deep learning research on sign language such as sign language recognition, sign language translation, and sign language segmentation for Korean sign language. There exist difficulties for deep learning research of sign language. First, it is difficult to recognize sign languages since they contain multiple modalities including hand movements, hand directions, and facial expressions. Second, it is the absence of training data to conduct deep learning research. Currently, KETI dataset is the only known dataset for Korean sign language for deep learning. Sign language datasets for deep learning research are classified into two categories: Isolated sign language and Continuous sign language. Although several foreign sign language datasets have been collected over time. they are also insufficient for deep learning research of sign language. Therefore, we attempted to collect a large-scale Korean sign language dataset and evaluate it using a baseline model named TSPNet which has the performance of SOTA in the field of sign language translation. The collected dataset consists of a total of 11,402 image and text. Our experimental result with the baseline model using the dataset shows BLEU-4 score 3.63, which would be used as a basic performance of a baseline model for Korean sign language dataset. We hope that our experience of collecting Korean sign language dataset helps facilitate further research directions on Korean sign language.

• Journal of the Korean Society of Radiology
• /
• v.18 no.1
• /
• pp.37-44
• /
• 2024

Even though MR can reveal excellent soft-tissue contrast and functional information, CT is also required for electron density information for accurate dose calculation in Radiotherapy. For the fusion of MRI and CT images in RT treatment planning workflow, patients are normally scanned on both MRI and CT imaging modalities. Recently deep-learning-based generations of CT images from MR images became possible owing to machine learning technology. This eliminated CT scanning work. This study implemented a CycleGan deep-learning-based CT image generation from MR images. Three CT generators whose learning is based on T1- , T2- , or T1-&T2-weighted MR images were created, respectively. We found that the T1-weighted MR image-based generator can generate better than other CT generators when T1-weighted MR images are input. In contrast, a T2-weighted MR image-based generator can generate better than other CT generators do when T2-weighted MR images are input. The results say that the CT generator from MR images is just outside the practical clinics and the specific weight MR image-based machine-learning generator can generate better CT images than other sequence MR image-based generators do.

• Korean Educational Research Journal
• /
• v.40 no.1
• /
• pp.17-42
• /
• 2019

Modality is the psychological attitude of the speaker, which is comprised by the sentences used in every language. Modality can be broadly categorized as perceptional modality and obligatory modality. This study summarizes the previous related literatures and theoretical branches of Korean linguistic studies. The study also proposes and classifies a modal concept on the Korean language, which is aimed at aiding Chinese people who are studying Korean. It further describes characteristics and expressions of modality in both the Chinese and Korean languages. This study aims to develop an effective teaching-learning program on the basis of the contrastive analysis between Korean language's modality, "-(으)ㄹ 수 있다," and the corresponding Chinese auxiliary verb, "能." Modality is a syntax item that reflects a speaker's subjective manner. There are many grammatical facets in Korean language books and teaching materials that are modal in nature. Further, modalities in Korean language are not only numerous but also have very rich meanings and functions. Based on the contrastive analysis, this study designs an effective teaching plan for Chinese people learning the Korean language. The designed system uses specific conversational occasions as the basis of learning, and it adapts the Korean language's modal system to classroom teaching. The system is expected to be effective during classroom teaching for demonstrating and learning modality in the Korean language.

• Archives of Plastic Surgery
• /
• v.45 no.5
• /
• pp.395-402
• /
• 2018

Increased emphasis on competency-based learning modules and widespread departure from traditional models of Halstedian apprenticeship have made surgical simulation an increasingly appealing component of medical education. Surgical simulators are available in numerous modalities, including virtual, synthetic, animal, and non-living models. The ideal surgical simulator would facilitate the acquisition and refinement of surgical skills prior to clinical application, by mimicking the size, color, texture, recoil, and environment of the operating room. Simulation training has proven helpful for advancing specific surgical skills and techniques, aiding in early and late resident learning curves. In this review, the current applications and potential benefits of incorporating simulation-based surgical training into residency curriculum are explored in depth, specifically in the context of plastic surgery. Despite the prevalence of simulation-based training models, there is a paucity of research on integration into resident programs. Current curriculums emphasize the ability to identify anatomical landmarks and procedural steps through virtual simulation. Although transfer of these skills to the operating room is promising, careful attention must be paid to mastery versus memorization. In the authors' opinions, curriculums should involve step-wise employment of diverse models in different stages of training to assess milestones. To date, the simulation of tactile experience that is reminiscent of real-time clinical scenarios remains challenging, and a sophisticated model has yet to be established.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.12 no.1
• /
• pp.392-412
• /
• 2018

Due to the semantic gap problem across different modalities, automatically retrieval from multimedia information still faces a main challenge. It is desirable to provide an effective joint model to bridge the gap and organize the relationships between them. In this work, we develop a deep image annotation and classification by fusing multi-modal semantic topics (DAC_mmst) model, which has the capacity for finding visual and non-visual topics by jointly modeling the image and loosely related text for deep image annotation while simultaneously learning and predicting the class label. More specifically, DAC_mmst depends on a non-parametric Bayesian model for estimating the best number of visual topics that can perfectly explain the image. To evaluate the effectiveness of our proposed algorithm, we collect a real-world dataset to conduct various experiments. The experimental results show our proposed DAC_mmst performs favorably in perplexity, image annotation and classification accuracy, comparing to several state-of-the-art methods.

• International Journal of Contents
• /
• v.18 no.3
• /
• pp.11-20
• /
• 2022

Human Emotion Recognition is an exciting topic that has been attracting many researchers for a lengthy time. In recent years, there has been an increasing interest in exploiting contextual information on emotion recognition. Some previous explorations in psychology show that emotional perception is impacted by facial expressions, as well as contextual information from the scene, such as human activities, interactions, and body poses. Those explorations initialize a trend in computer vision in exploring the critical role of contexts, by considering them as modalities to infer predicted emotion along with facial expressions. However, the contextual information has not been fully exploited. The scene emotion created by the surrounding environment, can shape how people perceive emotion. Besides, additive fusion in multimodal training fashion is not practical, because the contributions of each modality are not equal to the final prediction. The purpose of this paper was to contribute to this growing area of research, by exploring the effectiveness of the emotional scene gist in the input image, to infer the emotional state of the primary target. The emotional scene gist includes emotion, emotional feelings, and actions or events that directly trigger emotional reactions in the input image. We also present an attention-based fusion network, to combine multimodal features based on their impacts on the target emotional state. We demonstrate the effectiveness of the method, through a significant improvement on the EMOTIC dataset.

• The Journal of Information Systems
• /
• v.32 no.3
• /
• pp.1-17
• /
• 2023

Purpose The proliferation of data on the internet has created a need for innovative methods to analyze user satisfaction data. Traditional survey methods are becoming inadequate in dealing with the increasing volume and diversity of data, and new methods using unstructured internet data are being explored. While numerous comment-based user satisfaction studies have been conducted, only a few have explored user satisfaction through video and audio data. Multimodal sentiment analysis, which integrates multiple modalities, has gained attention due to its high accuracy and broad applicability. Design/methodology/approach This study uses multimodal sentiment analysis to analyze user satisfaction of iPhone and Samsung products through online videos. The research reveals that the combination model integrating multiple data sources showed the most superior performance. Findings The findings also indicate that price is a crucial factor influencing user satisfaction, and users tend to exhibit more positive emotions when content with a product's price. The study highlights the importance of considering multiple factors when evaluating user satisfaction and provides valuable insights into the effectiveness of different data sources for sentiment analysis of product reviews.

• Tuberculosis and Respiratory Diseases
• /
• v.52 no.1
• /
• pp.54-61
• /
• 2002

Background: Tracheal stenosis is an urgent but uncommon disease. Therefore, primary care clinicians have limited clinical experience. Animal models of a tracheal stenosis can be used conveniently for the learning, teaching, and developing new diagnostic and therapeutic modalities for tracheal stenosis. Recently, a canine model of a tracheal stenosis was developed using a Nd-YAG laser. To describe the methods and results of developed animal model, we performed this study. Methods : Six Mongrel dogs were generally anesthetized and the anterior 180 degree of tracheal cartilage of the animal was photo-coagulated using a Nd-YAG laser. The animals were bronchoscopically evaluated every week for 4 weeks and a pathologic evaluation was also made. Results : Two weeks after the laser coagulation, the trachea began to stenose and the stenosis progressed through 4 weeks. All animals suffered from shortness of breath, wheezing, and weight loss in the 3 weeks after the laser treatment, and two died of respiratory failure just before the fourth week. The gross pathologic findings showed the loss of cartilage and a dense fibrosis, which resulted in a fibrous stricture of the trachea. Microscopy also showed that the fibrous granulation tissue replaced destroyed cartilage. Conclusion : The canine model can assist in the understanding and development of new diagnostic and therapeutic modalities for tracheal stenosis.

• Archives of Plastic Surgery
• /
• v.47 no.1
• /
• pp.70-77
• /
• 2020

Background As the demand for cosmetic surgery continues to rise, plastic surgery programs and the training core curriculum have evolved to reflect these changes. This study aims to evaluate the perceived quality of current cosmetic surgery training in terms of case exposure and educational methods. Methods A 16-question survey was sent to graduates who completed their training at a U.S. plastic surgery training program in 2017. The survey assessed graduates' exposure to cosmetic surgery, teaching modalities employed and their overall perceived competence. Case complexity was characterized by the minimum number of cases needed by the graduate to feel confident in performing the procedure. Results There was a 25% response rate. The majority of respondents were residents (83%, n=92) and the remaining were fellows (17%, n=18). Almost three quarters of respondents were satisfied with their cosmetic training. Respondents rated virtual training as the most effective learning modality and observing attendings' patients/cases as least effective. Perceived competence was more closely aligned with core curriculum status than case complexity, i.e. graduates feel more prepared for core cosmetic procedures despite being more technically difficult than non-core procedures. Conclusions Despite the variability in cosmetic exposure during training, most plastic surgery graduates are satisfied with their aesthetic training. Incorporation of teaching modalities, such as virtual training, can increase case exposure and allow trainees more autonomy. The recommended core curriculum is adequately training plastic surgery graduates for common procedures and more specialized procedures should be consigned to aesthetic fellowship training.