• The KIPS Transactions:PartB
• /
• v.18B no.5
• /
• pp.305-314
• /
• 2011

Research into the visual perception of human emotion to recognize an intention has traditionally focused on emotions of facial expression. Recently researchers have turned to the more challenging field of emotional expressions through body posture or activity. Proposed work approaches recognition of basic emotional categories from body postures using neural model applied visual perception of neurophysiology. In keeping with information processing models of the visual cortex, this work constructs a biologically plausible hierarchy of neural detectors, which can discriminate 6 basic emotional states from static views of associated body postures of activity. The proposed model, which is tolerant to parameter variations, presents its possibility by evaluating against human test subjects on a set of body postures of activities.

• Journal of Internet Computing and Services
• /
• v.22 no.3
• /
• pp.9-16
• /
• 2021

This study aims to develop a human activity recognition (HAR) system as a Deep-Learning (DL) classification model, distinguishing various human activities. We solely rely on the signals from a wristband accelerometer worn by a person for the user's convenience. 3-axis sequential acceleration signal data are gathered within a predefined time-window-slice, and they are used as input to the classification system. We are particularly interested in developing a Deep-Learning model that can outperform conventional machine learning classification performance. A total of 13 activities based on the laboratory experiments' data are used for the initial performance comparison. We have improved classification performance using the Convolutional Neural Network (CNN) combined with an auto-encoder feature reduction and parameter tuning. With various publically available HAR datasets, we could also achieve significant improvement in HAR classification. Our CNN model is also compared against Recurrent-Neural-Network(RNN) with Long Short-Term Memory(LSTM) to demonstrate its superiority. Noticeably, our model could distinguish both general activities and near-identical activities such as sitting down on the chair and floor, with almost perfect classification accuracy.

• Journal of Korea Multimedia Society
• /
• v.23 no.8
• /
• pp.986-1005
• /
• 2020

In this paper, we propose AnoVid, an automated video annotation tool based on deep neural networks, that automatically generates various meta data for each scene or shot in a long drama video containing rich elements. To this end, a novel meta data schema for drama video is designed. Based on this schema, the AnoVid video annotation tool has a total of six deep neural network models for object detection, place recognition, time zone recognition, person recognition, activity detection, and description generation. Using these models, the AnoVid can generate rich video annotation data. In addition, AnoVid provides not only the ability to automatically generate a JSON-type video annotation data file, but also provides various visualization facilities to check the video content analysis results. Through experiments using a real drama video, "Misaeing", we show the practical effectiveness and performance of the proposed video annotation tool, AnoVid.

• Journal of Korea Multimedia Society
• /
• v.23 no.8
• /
• pp.977-985
• /
• 2020

Aiming at the problem of complex feature extraction and low accuracy in human action recognition, this paper proposed a network structure combining batch normalization algorithm with GoogLeNet network model. Applying Batch Normalization idea in the field of image classification to action recognition field, it improved the algorithm by normalizing the network input training sample by mini-batch. For convolutional network, RGB image was the spatial input, and stacked optical flows was the temporal input. Then, it fused the spatio-temporal networks to get the final action recognition result. It trained and evaluated the architecture on the standard video actions benchmarks of UCF101 and HMDB51, which achieved the accuracy of 93.42% and 67.82%. The results show that the improved convolutional neural network has a significant improvement in improving the recognition rate and has obvious advantages in action recognition.

• The Transactions of The Korean Institute of Electrical Engineers
• /
• v.59 no.4
• /
• pp.779-787
• /
• 2010

In this paper, We introduce electrical tracking generated from surface activity associated with flow of leakage current on insulator under wet and contaminated conditions and design electrical tracking pattern recognition system by using LabVIEW. We measure the leaking current of contaminated wire by using LabVIEW software and the NI-c-DAQ 9172 and NI-9239 hardware. As pattern recognition algorithm and optimization algorithm for electrical tracking system, neural networks, Radial Basis Function Neural Networks(RBFNNs) and particle swarm optimization are exploited. The designed electrical tracking recognition system consists of two parts such as the hardware part of electrical tracking generator, the NI-c-DAQ 9172 and NI-9239 hardware and the software part of LabVIEW block diagram, LabVIEW front panel and pattern recognition-related application software. The electrical tracking system decides whether electrical tracking generate or not on electrical wire.

• The Transactions of the Korean Institute of Electrical Engineers C
• /
• v.51 no.8
• /
• pp.361-367
• /
• 2002

Many failures in high voltage equipment are preceded by partial discharge activity. In this paper deals with the application of the high frequency partial discharge measurement technique in motorette. HFPD measurement is very effective method to detect the PD occurred in motorette which is the called name of test specimen for accelerating test of stator winding[1] In this study, CT type HFPD sensor is used to detect the partial discharges and a measured HFPD pattern is analyzed by fractal mathematics. The neural network algorithm is used to pattern recognition and ageing diagnosis. As a result of this study, the fractal dimensions are increased along to applied voltage and HFPD pattern recognition using neural network shown excellent recognition rate. Also, the ageing diagnosis of motorette has been Possible.

• Proceedings of the KIEE Conference
• /
• 2003.11c
• /
• pp.429-432
• /
• 2003

In this paper, a novel neural network approach is proposed for cleaning robot to complete coverage path planning with obstacle avoidance in stationary and dynamic environments. The dynamics of each neuron in the topologically organized neural network is characterized by a shunting equation derived from Hodgkin and Huxley's membrane equation. There are only local lateral connections among neurons. The robot path is autonomously generated from the dynamic activity landscape of the neural network and the previous robot location without any prior knowledge of the dynamic environment.

• Journal of Broadcast Engineering
• /
• v.27 no.7
• /
• pp.1021-1033
• /
• 2022

After the Versatile Video Coding (VVC)/H.266 standard was completed, the Joint Video Exploration Team (JVET) began to investigate new technologies that could significantly increase coding gain for the next generation video coding standard. One direction is to investigate signal processing based tools, while the other is to investigate Neural Network based technology. Neural Network based Video Coding (NNVC) has not been studied previously, and this is the first trial of such an approach in the standard group. After two years of research, JVET produced the first common software called Neural Compression Software (NCS) with two NN-based in-loop filtering tools at the 27th meeting and began to maintain NN-based technologies for the common experiment. The coding performances of the two filters in NCS-1.0 are shown to be 8.71% and 9.44% on average in a random access scenario, respectively. All the material related to NCS can be found in the repository of the JVET. In this paper, we provide a brief overview and review of the NNVC activity studied in JVET in order to provide trend and insight for the new direction of video coding standard.

• Journal of the Korean Society for Precision Engineering
• /
• v.22 no.7 s.172
• /
• pp.200-207
• /
• 2005

This study compares two different methods of measuring brain-BOLD activation. By comparing two different methods of measurement i.e., one method calculating the neural activation area (the number of activated voxels), while the other measured the neural activation intensity (the mean intensity of selected activated yokels), this study identified the more precise method of measuring brain activation which results from the completion of a visuospatial task. 16 right-handed male college students (mean age 23.2 years) participated in this study as subjects. Functional brain images were scanned on them using a 3T MRI single-shot EPI method. No correlation was found between the levels of cognitive performance and number of activated yokels in the activated brain areas. However, a significant correlation was found between the levels of cognitive performance and the mean intensity of selected activated yokels in the parietal, frontal, and other areas. In conclusion, the method of mean intensity was considered a better index of brain activity rather than the activated yokels measurement method.

• Physical Therapy Korea
• /
• v.24 no.2
• /
• pp.9-18
• /
• 2017

Background: Limitation of hamstring extensibility is often associated with various musculoskeletal problems such as alterations in posture and walking patterns. Thus, certain appropriate strategies need to be established for its management. Objects: The aim of this study was to compare the effects of the neural mobilization technique and static stretching exercises on popliteal angle and hamstring compliance in young women with short hamstring syndrome (SHS). Methods: Thirty-three women with SHS were randomly assigned to either group-1 ($n_1=17$) that underwent the neural mobilization technique or group-2 ($n_2=16$) that underwent the static stretching exercises. Outcome measures included the active popliteal angle (APA) and a hamstring's electromyographic (EMG) activity at a maximum popliteal angle of the baseline. Intervention for each group was performed for a total time of 3-min (6 sets of a 30-sec application). Results: There were significant interactions between time and group in the APA [group-1 (pre-test to post-test): $69.70{\pm}8.14^{\circ}$ to $74.14{\pm}8.07^{\circ}$ and group-2: $68.66{\pm}7.42^{\circ}$ to $70.52{\pm}7.92^{\circ}$] (F1,31=6.678, p=.015) and the EMG activity of the hamstring (group-1: $1.12{\pm}.30{\mu}N$ to $.69{\pm}.31{\mu}V$ and group-2: $1.19{\pm}.49{\mu}V$ to $1.13{\pm}.47{\mu}V$)(F1,31=6.678, p=.015). Between-group comparison revealed that the EMG activity of the hamstring was significantly different at post-test between the groups (p<.05). Furthermore, in within-group comparison, group-1 appeared to be significantly different for both variables between pre- and post-test (p<.05); however, group-2 showed significant difference in only the APA between pre- and post-test (p<.05). Conclusion: These findings suggest that the neural mobilization technique and static stretching exercises may be advantageous to improve hamstring compliance in young women with SHS, resulting in a more favorable outcome in the neural mobilization technique.