• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.36 no.2C
• /
• pp.78-85
• /
• 2011

In this paper we proposed an intelligent PTZ camera system which detects, classifies and tracks moving objects. If a moving object is detected, features are extracted for classification and then realtime tracking follows. We used GMM for detection followed by shadow removal. Legendre moment is used for classification. Without auto focusing, we can control the PTZ camera movement by using center points of the image and object's direction, distance and velocity. To implement the realtime system, we used TI DM6446 Davinci processor. Throughout the experiment, we obtained system's high performance in classification and tracking both at vehicle's normal and high speed motion.

• PNF and Movement
• /
• v.15 no.2
• /
• pp.133-140
• /
• 2017

Purpose: This study compares the effects of treadmill gait training accompanied by visual feedback and general treadmill gait training on the gait and balance ability of patients with chronic stroke. Methods: A total of 11 patients with chronic stroke were randomly divided into either the treadmill gait training accompanied by visual feedback group (six patients) or the general treadmill gait training group (five patients). The gait and balance ability of the two groups were measured before and after the interventions using the functional reach test, the Timed Up and Go (TUG) test, Berg's balance test, and the Biodex balance test. The treadmill gait training accompanied by visual feedback group performed the exercise under the supervision of a therapist after first being provided with a hat and a goal that was devised for the purpose of providing visual feedback information. The interventions were applied to the respective groups for four weeks. For the statistical analysis, we conducted a Mann-Whitney test to compare the results between the two groups. Additionally, the Wilcoxon test was used to compare the results from before and after the intervention in each group. Results: The treadmill gait training accompanied by visual feedback group showed a significant difference in terms of the functional reach test after the intervention when compared to the general treadmill gait training group (p<0.05). Although there was no significant difference, the treadmill gait training accompanied by visual feedback group showed a larger improvement in the TUG test, Berg's balance test, and the Biodex balance test than the general treadmill gait training group (p>0.05). Conclusion: The results of this study suggest that treadmill gait training accompanied by visual feedback can be used as a beneficial intervention scheme for the recovery of the gait and balance ability of patients with chronic stroke.

• Proceedings of the KSR Conference
• /
• 2010.06a
• /
• pp.2064-2068
• /
• 2010

In this paper, we introduced a sequential intelligent image analysis system(SIIAS). This system is implemented using PTZ camera with intelligent analysis algorithm and TI's Davinci DM6446. Enter, abandon, removal and cross functions are included in our system. These functions can be used individually or in combination for object monitoring and tracking. Sequential intelligent function processing is more efficient than the previous one by virtue of accurate observation, wide area monitoring and low cost.

• Proceedings of the Korean Information Science Society Conference
• /
• 2011.06c
• /
• pp.433-436
• /
• 2011

영상 통화 시스템을 설치한 장소에서 대두되는 문제점은 문서 유출 등의 보안문제와 사생활 침해가 발생하는 것이다. 본 논문에서는 실시간 영상 통화를 하면서 문서유출 및 보안사항에 위배되거나 사생활이 침해되는 영상이 발견된 즉시 영상 통화를 중단시켜 보안유지 및 사생활 보호를 위한 다빈치 기반 실시간 영상 통화 시스템을 개발하고자한다. 구현 된 시스템은 실시간 처리가 가능하고 영상통화 시 문서 유출 등 보안문제가 발생하는 영상을 검출하는 알고리즘을 개발하여 다빈치보드에 포팅하고 여러 가지 보안사항에 대한 실험에 적용하였다.

• Proceedings of the Korean Institute of Interior Design Conference
• /
• 2005.05a
• /
• pp.126-131
• /
• 2005

As our domestic Espresso Coffee Shops are in pursuit of professionalization and various coffee tastes, their sales will have been on increase. Now coffee market is getting more mature. Reflecting sensitivity of 20 to 30 generations, the domestic coffee shops provide them with best space for rest and conversation with stream of industralization and professionalization of the contemporary society. Based on such a situation, our research institute aims at studying how the interior designs put influence on marketing. Among the domestic franchise coffee shops, we will have researched Rosebud, Starbugs, Davinci which have more 50 chain stores throughout the country, This study aims to suggest a direction for interior design of brand coffee shop by analyzing brand coffee shops and style that customers prefer.

• Annual Conference on Human and Language Technology
• /
• 2023.10a
• /
• pp.171-176
• /
• 2023

최근 등장한 대규모 데이터로 사전학습된 자연어 생성 모델들은 대화 능력 및 코드 생성 태스크등에서 인상적인 성능을 보여주고 있어, 본 논문에서는 대형 언어 모델 (LLM)의 한국어 질문을 SQL 쿼리 (Text-to-SQL) 변환하는 성능을 평가하고자 한다. 먼저, 영어 Text-to-SQL 벤치마크 데이터셋을 활용하여 영어 질의문을 한국어 질의문으로 번역하여 한국어 Text-to-SQL 데이터셋으로 만들었다. 대형 생성형 모델 (GPT-3 davinci, GPT-3 turbo) 의 few-shot 세팅에서 성능 평가를 진행하며, fine-tuning 없이도 대형 언어 모델들의 경쟁력있는 한국어 Text-to-SQL 변환 성능을 확인한다. 또한, 에러 분석을 수행하여 한국어 문장을 데이터베이스 쿼리문으로 변환하는 과정에서 발생하는 다양한 문제와 프롬프트 기법을 활용한 가능한 해결책을 제시한다.

• Proceedings of the Korea Contents Association Conference
• /
• 2009.05a
• /
• pp.17-22
• /
• 2009

DaVinci processors are popular media processors for implementing embedded multimedia applications. They support dual core architecture: ARM9 core for video I/O handling as well as system management and peripheral handling, and DSP C64+ core for effective digital signal processing. In this paper, we propose our efforts for optimal implementation of object tracking algorithm in DaVinci-based smart camera which is being designed and implemented by our laboratory. The smart camera in this paper is supposed to support object detection, object tracking, object classification and detection of intrusion into surveillance regions and sending the detection event to remote clients using IP protocol. Object tracking algorithm is computationally expensive since it needs to process several procedures such as foreground mask extraction, foreground mask correction, connected component labeling, blob region calculation, object prediction, and etc. which require large amount of computation times. Thus, if it is not implemented optimally in Davinci-based processors, one cannot expect real-time performance of the smart camera.

• Journal of the Korean Society of Physical Medicine
• /
• v.15 no.4
• /
• pp.47-54
• /
• 2020

PURPOSE: This study compared two different techniques of treadmill gait training, and evaluated the outcomes on gait and respiratory function in patients with hemiplegic stroke. METHODS: This was a single-blinded, randomized, controlled, comparative study, enrolling a total of 21 stroke patients in a rehabilitation hospital. Subjects were randomly assigned to either the treadmill walking training group with gradual speed increase (GSI group, n = 10), or treadmill walking training group with random speed changing (RSC group, n = 11). All participants performed 60 min of comprehensive rehabilitation therapy (5 × / week for 6 weeks). In additional, each group received either GSI or RSC treadmill walking training for 20 min (5 × / week for 6 weeks, total 30 sessions). Gait and respiratory function were measured before and after the 6-week training. RESULTS: Both groups showed significant improvements in the 10-m walking test, 6-minute walking test, timed up and go test, forced vital capacity (FVC), forced expiratory volume in 1 second, and maximal voluntary ventilation after training (p < .05). The GSI group showed a significantly greater increase in the FVC than the RSC group (+14.8 L vs. +12.5 L, p < .05). CONCLUSION: Both training methods can be effective for improving the walking and respiratory functions of stroke patients. However, our results indicate that treadmill walking training with gradual speed increase might be a more effective method for improving the respiratory function (FVC) than treadmill walking training with random speed changing.

• The Journal of Korean Academy of Orthopedic Manual Physical Therapy
• /
• v.23 no.2
• /
• pp.33-44
• /
• 2017

Background: Lumbar stabilization exercise and gluteal muscle strengthening exercises are widely used to treat for lower back pain patient. The present study aimed to compare the effects of lumbar stabilization exercise and gluteal muscle strengthening exercises on chronic lower back pain with lumbar spine instability, with regard to pain intensity, disability, and psychosocial factors. Methods: Among 53 patients with chronic lower back pain, those with spine instability were selected using 5 examination tests. The selected 28 patients were randomly assigned to lumbar stabilization exercise group (LSE, n=15) and gluteal strengthening exercise group (GSE, n=13). Each group performed the corresponding exercise for 40 minutes, twice a week for 4 weeks. To analyze and compare the effects, pain intensity, the level of low back disability, and psychosocial factors were assessed before and after intervention. Results: There was significant difference in lower back pain intensity between the two groups before and after intervention. The change in low back disability was significant in the GSE group alone following intervention (p<.05), but no other significant difference was found between the groups. Among psychosocial factors, the changes in the fear-avoidance beliefs questionnaire (FABQ)-physical activity and FABQ-total were significant in the LSE group alone following intervention (p<.05). However, no significant difference were found in these factors between the two groups before and after intervention. Conclusions: LSE and GSE for lower back pain with lumbar spine instability showed no significant difference for pain intensity, physical disability, or psychosocial functioning.

• Journal of Digital Convergence
• /
• v.13 no.9
• /
• pp.225-233
• /
• 2015

The convergence social app manufacturing platform for both demander and developer in worldwide provides a collective collaborative development approach. This platform can bring enhanced productivity with exchange of views among participants. If the assessment is not fair that many participants will leave the collective collaborative project. The previous studies verified the reliability of the evaluation based on statistical techniques. Because the previous studies did not consider the task performance of participants, do not reflect the feature of project tasks. So, the contribution scores of the participants can be distorted. In this study, we suggest the method for evaluating the development contributions value. This is considered the task performance of participants and involved the method of equitable and consistency peer assessment.