• Journal of the Society of Disaster Information
• /
• v.19 no.1
• /
• pp.10-17
• /
• 2023

Purpose: In this paper, a distance estimation technique for indoor location estimation using IR-UWB is proposed and researched. We propose an algorithm that can increase the SNR lowered due to clutter or noise in an indoor environment. Method: In order to clutter suppression and detect weak signals, we analyze the existing studies of background remover, correlation, and singular vector decomposition techniques and propose an algorithm. Result: The proposed algorithm, the average error was 0.57m up to 11.5m, and the error were 0.49m from 6m to 11.5m. the average error rate was reduced by about 1m compared to the previous study. Conclusion: It can be used as a technique for detecting weak signals in clutter and noise environments for distance or location estimation, and can also be used as a human life detection technique to reduce damage to people in a disaster situation by using UWB radar which has highly transparent.

• Speech Sciences
• /
• v.11 no.2
• /
• pp.217-226
• /
• 2004

An automatic pronunciation correction system provides users with correction guidelines for each pronunciation error. For this purpose, we develop a speech recognition system which automatically classifies pronunciation errors when Koreans speak a foreign language. In this paper, we propose a machine scoring method for automatic assessment of pronunciation quality by the speech recognizer. Scores obtained from an expert human listener are used as the reference to evaluate the different machine scores and to provide targets when training some of algorithms. We use a log-likelihood score and a normalized log-likelihood score as machine scoring methods. Experimental results show that the normalized log-likelihood score had higher correlation with human scores than that obtained using the log-likelihood score.

• Journal of Multimedia Information System
• /
• v.3 no.2
• /
• pp.9-12
• /
• 2016

This paper is concerned with the effective and efficient identification of the gender and motion of humans. Tracking this nonverbal behavior is useful for providing clues about the interaction of different types of people and their exact motion. This system can also be useful for security in different places or for monitoring patients in hospital and many more applications. Here we describe a novel method of determining identity using machine learning with Microsoft Kinect. This method minimizes the fitting or overlapping error between an ellipsoid based skeleton.

• Proceedings of the Korea Contents Association Conference
• /
• 2009.05a
• /
• pp.839-842
• /
• 2009

It is used to differential image for moving object detection in general. But it is difficult to detect the accurate detection which uses differential image between frames. In this paper, the proposed method overcome the noise that is generated by camera, grabber card, or weather condition. It extract to moving big object such as human or vehicle. The proposed method process morphological filtering and binary for the image with noise, reduce error. We are expect to apply to a real-time moving object detection system at fog condition, pass the limit of the object detection method using the differential image.

• The Journal of Korea Robotics Society
• /
• v.16 no.4
• /
• pp.336-344
• /
• 2021

This paper presents a robust deep learning-based human tracking framework in crowded environments. For practical human tracking applications, a target must be robustly tracked even in undetected or overcrowded situations. The proposed framework consists of two parts: robust deep learning-based human detection and tracking while recognizing the aforementioned situations. In the former part, target candidates are detected using Detectron2, which is one of the powerful deep learning tools, and their weights are computed and assigned. Subsequently, a candidate with the highest weight is extracted and is utilized to track the target human using a Kalman filter. If the bounding boxes of the extracted candidate and another candidate are overlapped, it is regarded as a crowded situation. In this situation, the center information of the extracted candidate is compensated using the state estimated prior to the crowded situation. When candidates are not detected from Detectron2, it means that the target is completely occluded and the next state of the target is estimated using the Kalman prediction step only. In two experiments, people wearing the same color clothes and having a similar height roam around the given place by overlapping one another. The average error of the proposed framework was measured and compared with one of the conventional approaches. In the error result, the proposed framework showed its robustness in the crowded environments.

• Korean Journal of Remote Sensing
• /
• v.34 no.2_2
• /
• pp.327-338
• /
• 2018

The possibility of ship detection monitoring at operational level using KOMPSAT-5 Synthetic Aperture Radar (SAR) and Automatic Identification System (AIS) data is investigated. For the analysis, the KOMPSAT-5 SLC images, which are collected from the west coast of Shinjin port and the northern coast of Jeju port are used along with portable AIS data from near the coast. The ship detection algorithm based on HVAS (Human Visual Attention System) was applied, which has significant advantages in terms of detection speed and accuracy compared to the commonly used CFAR (Constant False Alarm Rate). As a result of the integrated analysis, the ship detection from KOMPSAT-5 and AIS were generally consistent except for small vessels. Some ships detected in KOMPSAT-5 but not in AIS are due to the data absence from AIS, while it is clearly visible in KOMPSAT-5. Meanwhile, SAR imagery also has some false alarms due to ship wakes, ghost effect, and DEM error (or satellite orbit error) during object masking in land. Improving the developed ship detection algorithm and collecting reliable AIS data will contribute for building wide integrated surveillance system of marine territory at operational level.

• Journal of Intelligence and Information Systems
• /
• v.29 no.3
• /
• pp.267-286
• /
• 2023

The conversational agents such as AI speakers utilize voice conversation for human-computer interaction. Voice recognition errors often occur in conversational situations. Recognition errors in user utterance records can be categorized into two types. The first type is misrecognition errors, where the agent fails to recognize the user's speech entirely. The second type is misinterpretation errors, where the user's speech is recognized and services are provided, but the interpretation differs from the user's intention. Among these, misinterpretation errors require separate error detection as they are recorded as successful service interactions. In this study, various text separation methods were applied to detect misinterpretation. For each of these text separation methods, the similarity of consecutive speech pairs using word embedding and document embedding techniques, which convert words and documents into vectors. This approach goes beyond simple word-based similarity calculation to explore a new method for detecting misinterpretation errors. The research method involved utilizing real user utterance records to train and develop a detection model by applying patterns of misinterpretation error causes. The results revealed that the most significant analysis result was obtained through initial consonant extraction for detecting misinterpretation errors caused by the use of unregistered neologisms. Through comparison with other separation methods, different error types could be observed. This study has two main implications. First, for misinterpretation errors that are difficult to detect due to lack of recognition, the study proposed diverse text separation methods and found a novel method that improved performance remarkably. Second, if this is applied to conversational agents or voice recognition services requiring neologism detection, patterns of errors occurring from the voice recognition stage can be specified. The study proposed and verified that even if not categorized as errors, services can be provided according to user-desired results.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2017.10a
• /
• pp.465-469
• /
• 2017

In this paper we propose a human body tracking method for application of smartphones. The conventional human body tracking method is divided into a sensor-based method and a vision-based method. The sensor-based methods have a weakness in that tracking accuracy is low due to cumulative error of position information. The vision-based method has no cumulative error, but it requires reduction of the computational complexity for application of smartphone. In this paper we use the improved HOG algorithm as a human body tracking method for application of smartphone. The improved HOG algorithm is implemented through downsampling and frame sampling. Gaussian pyramid is applied for downsampling, and uniform sampling is applied for frame sampling. We measured the proposed algorithm on two devices, four resolutions, and four frame sampling intervals. We derive the best detection rate among downsampling and frame sampling parameters that can be applied in realtime.

• The Journal of Korean Society for Radiation Therapy
• /
• v.21 no.1
• /
• pp.33-39
• /
• 2009

Purpose: The aim of this study is to compare patient's body posture and its position at the time of simulation with one at the treatment room using On-board Imaging (OBI) and CT (CBCT). The detected offsets are compared with position errors of Rando Phantom that are practically applied. After that, Rando Phantom's position is selected by moving couch based on detected deviations. In addition, the errors between real measured values of Rando Phantom position and theoretical ones is compared. And we will evaluate target position's accuracy of KV X-ray imaging's 2D and CBCT's 3D one. Materials and Methods: Using the Rando Phantom (Alderson Research Laboratories Inc. Stanford. CT, USA) which simulated human body's internal structure, we will set up Rando Phantom on the treatment couch after implementing simulation and RTP according to the same ways as the real radioactive treatment. We tested Rando Phantom that are assumed to have accurate position with different 3 methods. We measured setup errors on the axis of X, Y and Z, and got mean standard deviation errors by repeating tests 10 times on each tests. Results: The difference between mean detection error and standard deviation are as follows; lateral 0.4+/-0.3 mm, longitudinal 0.6+/-0.5 mm, vertical 0.4+/-0.2 mm which all within 0~10 mm. The couch shift variable after positioning that are comparable to residual errors are 0.3+/-0.1, 0.5+/-0.1, and 0.3+/-0.1 mm. The mean detection errors by longitudinal shift between 20~40 mm are 0.4+/-0.3 in lateral, 0.6+/-0.5 in longitudinal, 0.5+/-0.3 in vertical direction. The detection errors are all within range of 0.3~0.5 mm. Residual errors are within 0.2~0.5 mm. Each values are mean values based on 3 tests. Conclusion: Phantom is based on treatment couch shift and error within the average 5mm can be gained by the diminution detected by image registration based on OBI and CBCT. Therefore, the selection of target position which depends on OBI and CBCT could be considered as useful.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.16 no.2
• /
• pp.49-54
• /
• 2016

In this paper, we propose an Arduino-based effective home security monitoring system. Proposed home security monitoring system consists of arduino which is inexpensive main processor, ultrasonic sensor and human body detection sensor to detect whether someone breaks into home. Data from ultrasonic sensor and human body detection sensor are transmitted to web server via ethernet shield connected to arduino. Web server checks whether someone breaks into home by using stored data from ultrasonic sensor and human body detection sensor. Snapshot is photographed via webcam connected by using JQuery. Photographed snapshot is stored in web server as image file. A user can monitor in web or smart device environment by using HTML5, CSS and Canvas. When examining efficiency of proposed home security monitoring system, it was found that proposed system is easier to be made than existing home security system and is cost effective by using arduino and is efficient and convenient and stable as it enables a user to handle an error in person and it uses reliable data.