• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.26 no.5
• /
• pp.668-674
• /
• 2022

Machine learning techniques using visual data have high usability in fields of industry and service such as scene recognition, fault detection, security and user analysis. Among these, user analysis through the videos from CCTV is one of the practical way of using vision data. Also, many studies about lightweight artificial neural network have been published to increase high usability for mobile and embedded environment so far. In this study, we propose the network combining the object detection and classification for mobile graphic processing unit. This network detects pedestrian and face, classifies age and gender from detected face. Proposed network is constructed based on MobileNet, YOLOv2 and skip connection. Both detection and classification models are trained individually and combined as 2-stage structure. Also, attention mechanism is used to improve detection and classification ability. Nvidia Jetson Nano is used to run and evaluate the proposed system.

• Journal of Platform Technology
• /
• v.11 no.3
• /
• pp.56-67
• /
• 2023

Recently, de-identification of personal information, which has been a long-cherished desire of the data-based industry, was revised and specified in August 2020. It became the foundation for activating data called crude oil[2] in the fourth industrial era in the industrial field. However, some people are concerned about the infringement of the basic rights of the data subject[3]. Accordingly, a development study was conducted on the Batch De-Identification Tool, a personal information de-identification automation tool. In this study, first, we developed an image labeling tool to label human faces (eyes, nose, mouth) and car license plates of various resolutions to build data for training. Second, an object recognition model was trained to run the object recognition module to perform de-identification of personal information. The automated personal information de-identification tool developed as a result of this research shows the possibility of proactively eliminating privacy violations through online services. These results suggest possibilities for data-based industries to maximize the value of data while balancing privacy and utilization.

• Smart Media Journal
• /
• v.12 no.7
• /
• pp.59-67
• /
• 2023

Recently, research on behavior analysis tracking human posture and movement has been actively conducted. In particular, OpenPose, an open-source software developed by CMU in 2017, is a representative method for estimating human appearance and behavior. OpenPose can detect and estimate various body parts of a person, such as height, face, and hands in real-time, making it applicable to various fields such as smart healthcare, exercise training, security systems, and medical fields. In this paper, we propose a method for classifying four exercise movements - Squat, Walk, Wave, and Fall-down - which are most commonly performed by users in the gym, using OpenPose-based deep learning models, DNN and CNN. The training data is collected by capturing the user's movements through recorded videos and real-time camera captures. The collected dataset undergoes preprocessing using OpenPose. The preprocessed dataset is then used to train the proposed DNN and CNN models for exercise movement classification. The performance errors of the proposed models are evaluated using MSE, RMSE, and MAE. The performance evaluation results showed that the proposed DNN model outperformed the proposed CNN model.

• Journal of Creative Information Culture
• /
• v.4 no.2
• /
• pp.95-101
• /
• 2018

As a result of introducing robot assisted learning which utilizes social robots or telepresence robots in language learning or special education, research on technology acceptance model for robot-assisted learning is also being conducted. The unified theory of acceptance and use of technology (UTAUT) model of intelligent robot has been studied, but of tele-operated robot is insufficient. The purpose of this paper is to estimate the UTAUT model by pre-service teachers who experienced telepresence robot-assisted learning that can be done in future school. It is found that the estimated UTAUT model consists of more concise factors than social robots, and the importance of perceived enjoyment is higher. In other words, the pre-service teachers showed significant acceptance of tele-operated robots with enhanced enjoyment composed of its mobility, communication, and touchable appearance of the face and body.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.05a
• /
• pp.329-332
• /
• 2022

This study aims to identify problems with existing e-learning content and non-face-to-face class methods, improve students' concentration, improve class achievement and educational effectiveness, and propose an artificial intelligence class system design using a web server. By using the function of face and eye tracking using OpenCV to identify attendance and concentration, and by inducing feedback through voice or message to questions asked by the instructor in the middle of class, learners relieve boredom caused by online classes and test by runner If the score is not reached, we propose an artificial intelligence education program system design that can bridge the academic gap and improve academic achievement by providing educational materials and videos for the wrong problem.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2022.10a
• /
• pp.280-283
• /
• 2022

This study aims to identify problems with existing e-learning content and non-face-to-face class methods, improve students' concentration, improve class achievement and educational effectiveness, and propose an artificial intelligence class system design using a web server. By using the function of face and eye tracking using OpenCV to identify attendance and concentration, and by inducing feedback through voice or message to questions asked by the instructor in the middle of class, learners relieve boredom caused by online classes and test by runner If the score is not reached, we propose an artificial intelligence education program system design that can bridge the academic gap and improve academic achievement by providing educational materials and videos for the wrong problem.

• Imaging Science in Dentistry
• /
• v.36 no.1
• /
• pp.49-54
• /
• 2006

Purpose : To evaluate accuracy and reliability of program to measure facial soft tissue thickness using 3D computed tomographic images by comparing with direct measurement. Materials and Methods : One cadaver was scanned with a Helical CT with 3 mm slice thickness and 3 mm/sec table speed. The acquired data was reconstructed with 1.5 mm reconstruction interval and the images were transferred to a personal computer. The facial soft tissue thickness were measured using a program developed newly in 3D image. For direct measurement, the cadaver was cut with a bone cutter and then a ruler was placed above the cut side. The procedure was followed by taking pictures of the facial soft tissues with a high-resolution digital camera. Then the measurements were done in the photographic images and repeated for ten times. A repeated measure analysis of variance was adopted to compare and analyze the measurements resulting from the two different methods. Comparison according to the areas was analyzed by Mann-Whitney test. Results : There were no statistically significant differences between the direct measurements and those using the 3D images (p>0.05). There were statistical differences in the measurements on 17 points but all the points except 2 points showed a mean difference of 0.5 mm or less. Conclusion : The developed software program to measure the facial soft tissue thickness using 3D images was so accurate that it allows to measure facial soft tissues thickness more easily in forensic science and anthropology.

• Journal of the Institute of Electronics Engineers of Korea SD
• /
• v.45 no.6
• /
• pp.49-59
• /
• 2008

SIFT(Scale Invariant Feature Transform) is an algorithm to extract vectors at pixels around keypoints, in which the pixel colors are very different from neighbors, such as vortices and edges of an object. The SIFT algorithm is being actively researched for various image processing applications including 3-D image constructions, and its most computation-intensive stage is a keypoint localization. In this paper, we develope a fixed-point model of the keypoint localization and propose its efficient hardware architecture for embedded applications. The bit-length of key variables are determined based on two performance measures: localization accuracy and error rate. Comparing with the original algorithm (implemented in Matlab), the accuracy and error rate of the proposed fixed point model are 93.57% and 2.72% respectively. In addition, we found that most of missing keypoints appeared at the edges of an object which are not very important in the case of keypoints matching. We estimate that the hardware implementation will give processing speed of $10{\sim}15\;frame/sec$, while its fixed point implementation on Pentium Core2Duo (2.13 GHz) and ARM9 (400 MHz) takes 10 seconds and one hour each to process a frame.

• Journal of the Institute of Electronics and Information Engineers
• /
• v.53 no.4
• /
• pp.159-166
• /
• 2016

In this paper, an active object tracking system basing on the pan/tilt-embedded stereo camera system is suggested and implemented. In the proposed system, once the face area of a target is detected from the input stereo image by using a YCbCr color model and phase-type correlation scheme and then, using this data as well as the geometric information of the tracking system, the distance and 3D information of the target are effectively extracted in real-time. Basing on these extracted data the pan/tilted-embedded stereo camera system is adaptively controlled and as a result, the proposed system can track the target adaptively under the various circumstance of the target. From some experiments using 480 frames of the test input stereo image, it is analyzed that a standard variation between the measured and computed the estimated target's height and an error ratio between the measured and computed 3D coordinate values of the target is also kept to be very low value of 1.03 and 1.18% on average, respectively. From these good experimental results a possibility of implementing a new real-time intelligent stereo target tracking and surveillance system using the proposed scheme is finally suggested.

• Science of Emotion and Sensibility
• /
• v.8 no.2
• /
• pp.161-167
• /
• 2005

The study is to examine how facial EMG responses change when children experience a positive emotion(happiness) and a negative emotion(fear). It is to prove that the positive emotion(happiness) could be distinguishable from the negative emotion(fear) by the EMG responses. Audiovisual film clips were used for evoking the positive emotion(happiness) and the negative emotion(fear). 47 children (11-13 years old, 23 boys and 24 girls) participated in the study Facial EMG (right corrugator and orbicularis oris) was measured while children were experiencing the positive or negative emotion. Emotional assessment scale was used for measuring children's psychological responses. It showed more than $85\%$ appropriateness and 3.15, 4.04 effectiveness (5 scale) for happiness and fear, respectively. Facial EMG responses were significantly different between a resting state and a emotional state both in happiness and in fear (p<001). Result suggests that each emotion was distinguishable by corrugator and orbicularis oris responses. Specifically, corrugator was more activated in the positive emotion(happiness) than in the negative emotion(fear), whereas orbicularis oris was more activated in the negative emotion(fear) than in the positive emotion(fear).