• Journal of the Korean Society for information Management
• /
• v.34 no.1
• /
• pp.219-239
• /
• 2017

As IT technology develops rapidly and the personal dissemination of smart devices increases, video material is especially used as a medium of information transmission among audiovisual materials. Video as an information service content has become an indispensable element, and it has been used in various ways such as unidirectional delivery through TV, interactive service through the Internet, and audiovisual library borrowing. Especially, in the Internet environment, the information provider tries to reduce the effort and cost for the processing of the provided information in view of the video service through the smart device. In addition, users want to utilize only the desired parts because of the burden on excessive network usage, time and space constraints. Therefore, it is necessary to enhance the usability of the video by automatically classifying, summarizing, and indexing similar parts of the contents. In this paper, we propose a method of automatically segmenting the shots that make up videos by analyzing the contents and characteristics of language education videos and indexing the detailed contents information of the linguistic videos by combining visual features. The accuracy of the semantic based shot segmentation is high, and it can be effectively applied to the summary service of language education videos.

• Investigative Magnetic Resonance Imaging
• /
• v.11 no.1
• /
• pp.39-48
• /
• 2007

Purpose : To determine the diagnostic accuracy of four different sequences : moderately T2 weighted, two heavily T2-weighted single shot turbo spin-echo sequence and breath-hold axial-2D balanced turbo field-echo sequence(bFFE) for characterization of focal lesions. Materials and Methods : During the 3-month period between June and August 2005, seventy-six patients were proved to have ninety-three focal hepatic lesions on MR imaging. The patients consisted of 49 men and 27 women (age range, 15-75 years; mean age, 56.23 years). All MR images were acquired on a 1.5-T MR using the following sequences: 1. A breath-hold axial T2-weighted single shot turbo spin-echo sequence, 2. a breath-hold axial-2D balanced turbo field-echo sequence. Two radiologists performed quantitative analysis. Another radiologist measured the lesion-to-liver contrast-to-noise ratio at the region-of-interest in the four sequences. Results : There was no significant difference in inter-observer variability between the four sequences. The accuracy for both cyst and malignancy of moderate T2 weighted MRI (echo time: 80 msec) was also highest. There was significant difference for lesion characterization between moderate T2 weighted MRI and balanced steady state procession (p-value: 0.004) in the second reader. For longer echo time, the CNR of cystic lesions were markedly increased in comparison to lesions of other component. Conclusion : The accuracy and inter-observer variability of single shot turbo spin echo T2 weighted sequence was higher than bFFE. Although there was no statically significant difference, moderate T2 weighted MRI (echo time: 80 msec) was more accurate than heavily T2 weighted sequence (echo time: 300 msec). If the results for lesion characterization is equivocal in TE 80, the addition of heavily T2 weighted MRI (echo time: 180 msec) can be helpful.

• Journal of Biomedical Engineering Research
• /
• v.24 no.4
• /
• pp.267-273
• /
• 2003

The present study attempted to observe what changes the supply of highly concentrated (30%) oxygen cause to people's ability of visuospatial cognition, compared to air of normal oxygen concentration (21%). This study sampled eight male university students (the average age : 23.5) as subjects for functional Magnetic Resonance Imaging (MRI) study It also developed equipment that supplies 21% and 30% oxygen) at a constant rate of 8L/min. Two questionnaires containing 20 questions were developed to measure the ability of visuospatial cognition, and accuracy was calculated from the result of task performance. The experiment paradigm consisted of the run conducting tasks at 30%'s concentration of oxygen and another run at 21%'s concentration of oxygen. Each run was composed of four blocks and each block included eight control tasks and five visuospatial taks. 3T MRI was used and fMRI was obtained through the single-shot EPI method. The activation in the occipital-associated area, bilateral superior parietal lobes, bilateral inferior parietal lobes. bilateral precuneus, bilateral postcentral gyri, bilateral middle frontal gyri, bilateral inferior frontal gyri, bilateral medial frontal gyri, bilateral superior frontal gyri, bilateral cingulate gyri was significantly increased at the 30%'s concentration of oxygen rather than 21%'s. Furthermore, the result of task performance showed the accuracy increased at 30%'s concentration of oxygen rather than 21%'s. From the result of this study, it is concluded that the supply of highly concentrated oxygen has a positive effect on the ability of visuospatial cognition.

• ETRI Journal
• /
• v.40 no.4
• /
• pp.421-434
• /
• 2018

The application of deep neural networks (DNNs) to connect the world with cyber physical systems (CPSs) has attracted much attention. However, DNNs require a large amount of memory and computational cost, which hinders their use in the relatively low-end smart devices that are widely used in CPSs. In this paper, we aim to determine whether DNNs can be efficiently deployed and operated in low-end smart devices. To do this, we develop a method to reduce the memory requirement of DNNs and increase the inference speed, while maintaining the performance (for example, accuracy) close to the original level. The parameters of DNNs are decomposed using a hybrid of canonical polyadic-singular value decomposition, approximated using a tensor power method, and fine-tuned by performing iterative one-shot hybrid fine-tuning to recover from a decreased accuracy. In this study, we evaluate our method on frequently used networks. We also present results from extensive experiments on the effects of several fine-tuning methods, the importance of iterative fine-tuning, and decomposition techniques. We demonstrate the effectiveness of the proposed method by deploying compressed networks in smartphones.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.13 no.10
• /
• pp.5095-5111
• /
• 2019

A multi-rotor Unmanned Aerial Vehicle (UAV) system is developed to solve the manhole cover detection problem for the infrastructure maintenance in the suburbs of big city. The visible light sensor is employed to collect the ground image data and a series of image processing and machine learning methods are used to detect the manhole cover. First, the image enhancement technique is employed to improve the imaging effect of visible light camera. An imaging environment perception method is used to increase the computation robustness: the blind Image Quality Evaluation Metrics (IQEMs) are used to percept the imaging environment and select the images which have a high imaging definition for the following computation. Because of its excellent processing effect the adaptive Multiple Scale Retinex (MSR) is used to enhance the imaging quality. Second, the Single Shot multi-box Detector (SSD) method is utilized to identify the manhole cover for its stable processing effect. Third, the spatial coordinate of manhole cover is also estimated from the ground image. The practical applications have verified the outdoor environment adaptability of proposed algorithm and the target detection correctness of proposed system. The detection accuracy can reach 99% and the positioning accuracy is about 0.7 meters.

• 한국HCI학회:학술대회논문집
• /
• 2006.02a
• /
• pp.711-716
• /
• 2006

Since the introduction of digital camera to the mass market, the number of digital photos owned by an individual is growing at an alarming rate. This phenomenon naturally leads to the issues of difficulties while searching and browsing in the personal digital photo archive. Traditional approach typically involves content-based image retrieval using computer vision algorithms. However, due to the performance limitations of these algorithms, at least on the casual digital photos taken by non-professional photographers, more recent approaches are centered on time-based clustering algorithms, analyzing the shot times of photos. These time-based clustering algorithms are based on the insight that when these photos are clustered according to the shot-time similarity, we have "event clusters" that will help the user browse through her photo archive. It is also reported that one of the remaining problems with the time-based approach is that people perceive events in different scales. In this paper, we present an adaptive time-based clustering algorithm that exploits the usage history of digital photos in order to infer the user's preference on the event granularity. Experiments show significant performance improvements in the clustering accuracy.

• The Journal of Korean Institute of Next Generation Computing
• /
• v.15 no.5
• /
• pp.53-63
• /
• 2019

To provide a hand gesture interface through deep learning in mobile environments, research on the light-weighting of networks is essential for high recognition rates while at the same time preventing degradation of execution speed. This paper proposes a method of real-time recognition of written characters in the air using a finger on mobile devices through the light-weighting of deep-learning model. Based on the SSD (Single Shot Detector), which is an object detection model that utilizes MobileNet as a feature extractor, it detects index finger and generates a result text image by following fingertip path. Then, the image is sent to the server to recognize the characters based on the learned OCR model. To verify our method, 12 users tested 1,000 words using a GALAXY S10+ and recognized their finger with an average accuracy of 88.6%, indicating that recognized text was printed within 124 ms and could be used in real-time. Results of this research can be used to send simple text messages, memos, and air signatures using a finger in mobile environments.

• Transactions on Semiconductor Engineering
• /
• v.2 no.1
• /
• pp.17-20
• /
• 2024

This paper presents a novel lightweight object detection model tailored for low-powered edge devices, addressing the limitations of traditional resource-intensive computer vision models. Our proposed detector, inspired by the Single Shot Detector (SSD), employs a compact yet robust network design. Crucially, it integrates an 'enhancer block' that significantly boosts its efficiency in detecting smaller objects. The model comprises two primary components: the Light_Block for efficient feature extraction using Depth-wise and Pointwise Convolution layers, and the Enhancer_Block for enhanced detection of tiny objects. Trained from scratch on the Udacity Annotated Dataset with image dimensions of 300x480, our model eschews the need for pre-trained classification weights. Weighing only 5.5MB with approximately 0.43M parameters, our detector achieved a mean average precision (mAP) of 27.7% and processed at 140 FPS, outperforming conventional models in both precision and efficiency. This research underscores the potential of lightweight designs in advancing object detection for edge devices without compromising accuracy.

• Proceedings of the Korean Society for Emotion and Sensibility Conference
• /
• 2003.11a
• /
• pp.81-83
• /
• 2003

The present study attempted to observe what changes the supply of highly concentrated(30％) oxygen cause to people's ability and cerebrum lateralization of verbal cognition, compared to air of normal oxygen concentration(21％). The experiment consisted of two runs, one for verbal cognition test with normal air(21％ of oxygen) and for verbal cognition test with more oxygen in the air(30％ of oxygen). Functional brain images were taken form 3T MRI using the single-shot EPI method. There were more activations observed at the occipital, parietal, temporal, and frontal lobes, but there were no changes in cerebrum lateralization with 30％ oxygen administration. The result of task performance showed the accuracy increased at 30％'s concentration of oxygen rather than 21％'s. It is concluded that the positive effect on the verbal cognitive performance level by the highly concentrated oxygen administration was due to changeless increase of left and right cerebrum activation.

• Journal of the Korea Academia-Industrial cooperation Society
• /
• v.12 no.2
• /
• pp.581-587
• /
• 2011

Reliability estimation methods using attribute data are widely used in reliability evaluation of various systems such as nuclear energy plants, food and drug, and space launch vehicles. This paper shows sensitivity analysis and comparison results of reliability estimation methods including a parametric estimation method in open literature with respect to both sample size and sampling points of time. And ways to improve accuracy of each reliability estimation method were proposed from the sensitivity analysis results.