• Journal of Intelligence and Information Systems
• /
• v.25 no.1
• /
• pp.163-177
• /
• 2019

As smartphones are getting widely used, human activity recognition (HAR) tasks for recognizing personal activities of smartphone users with multimodal data have been actively studied recently. The research area is expanding from the recognition of the simple body movement of an individual user to the recognition of low-level behavior and high-level behavior. However, HAR tasks for recognizing interaction behavior with other people, such as whether the user is accompanying or communicating with someone else, have gotten less attention so far. And previous research for recognizing interaction behavior has usually depended on audio, Bluetooth, and Wi-Fi sensors, which are vulnerable to privacy issues and require much time to collect enough data. Whereas physical sensors including accelerometer, magnetic field and gyroscope sensors are less vulnerable to privacy issues and can collect a large amount of data within a short time. In this paper, a method for detecting accompanying status based on deep learning model by only using multimodal physical sensor data, such as an accelerometer, magnetic field and gyroscope, was proposed. The accompanying status was defined as a redefinition of a part of the user interaction behavior, including whether the user is accompanying with an acquaintance at a close distance and the user is actively communicating with the acquaintance. A framework based on convolutional neural networks (CNN) and long short-term memory (LSTM) recurrent networks for classifying accompanying and conversation was proposed. First, a data preprocessing method which consists of time synchronization of multimodal data from different physical sensors, data normalization and sequence data generation was introduced. We applied the nearest interpolation to synchronize the time of collected data from different sensors. Normalization was performed for each x, y, z axis value of the sensor data, and the sequence data was generated according to the sliding window method. Then, the sequence data became the input for CNN, where feature maps representing local dependencies of the original sequence are extracted. The CNN consisted of 3 convolutional layers and did not have a pooling layer to maintain the temporal information of the sequence data. Next, LSTM recurrent networks received the feature maps, learned long-term dependencies from them and extracted features. The LSTM recurrent networks consisted of two layers, each with 128 cells. Finally, the extracted features were used for classification by softmax classifier. The loss function of the model was cross entropy function and the weights of the model were randomly initialized on a normal distribution with an average of 0 and a standard deviation of 0.1. The model was trained using adaptive moment estimation (ADAM) optimization algorithm and the mini batch size was set to 128. We applied dropout to input values of the LSTM recurrent networks to prevent overfitting. The initial learning rate was set to 0.001, and it decreased exponentially by 0.99 at the end of each epoch training. An Android smartphone application was developed and released to collect data. We collected smartphone data for a total of 18 subjects. Using the data, the model classified accompanying and conversation by 98.74% and 98.83% accuracy each. Both the F1 score and accuracy of the model were higher than the F1 score and accuracy of the majority vote classifier, support vector machine, and deep recurrent neural network. In the future research, we will focus on more rigorous multimodal sensor data synchronization methods that minimize the time stamp differences. In addition, we will further study transfer learning method that enables transfer of trained models tailored to the training data to the evaluation data that follows a different distribution. It is expected that a model capable of exhibiting robust recognition performance against changes in data that is not considered in the model learning stage will be obtained.

• The Journal of Engineering Geology
• /
• v.13 no.3
• /
• pp.293-308
• /
• 2003

New results about the crustal structure down to a depth of 60 km beneath North Korea were obtained using the seismic tomography method. About 1013 P- and S-wave travel times from local earthquakes recorded by the Korean stations and the vicinity were used in the research. All earthquakes were relocated on the basis of an algorithm proposed in this study. Parameterization of the velocity structure is realized with a set of nodes distributed in the study volume according to the ray density. 120 nodes located at four depth levels were used to obtain the resulting P- and S-wave velocity structures. As a result, it is found that P- and S-wave velocity anomalies of the Rangnim Massif at depth of 8 km are high and low, respectively, whereas those of the Pyongnam Basin are low up to 24 km. It indicates that the Rangnim Massif contains Archean-early Lower Proterozoic Massif foldings with many faults and fractures which may be saturated with underground water and/or hot springs. On the other hand, the Pyongyang-Sariwon in the Pyongnam Basin is an intraplatform depression which was filled with sediments for the motion of the Upper Proterozoic, Silurian and Upper Paleozoic, and Lower Mesozoic origin. In particular, the high P- and S-wave velocity anomalies are observed at depth of 8, 16, and 24 km beneath Mt. Backdu, indicating that they may be the shallow conduits of the solidified magma bodies, while the low P-and S-wave velocity anomalies at depth of 38 km must be related with the magma chamber of low velocity bodies with partial melting. We also found the Moho discontinuities beneath the Origin Basin including Sari won to be about 55 km deep, whereas those of Mt. Backdu is found to be about 38 km. The high ratio of P-wave velocity/S-wave velocity at Moho suggests that there must be a partial melting body near the boundary of the crust and mantle. Consequently we may well consider Mt. Backdu as a dormant volcano which is holding the intermediate magma chamber near the Moho discontinuity. This study also brought interesting and important findings that there exist some materials with very high P- and S-wave velocity annomoalies at depth of about 40 km near Mt. Myohyang area at the edge of the Rangnim Massif shield.

• The Korean Journal of Nuclear Medicine
• /
• v.39 no.6
• /
• pp.413-420
• /
• 2005

Purpose: Neuroreceptor PET studies require 60-120 minutes to complete and head motion of the subject during the PET scan increases the uncertainty in measured activity. In this study, we investigated the effects of the data-driven head mutton correction on the evaluation of endogenous dopamine release (DAR) in the striatum during the motor task which might have caused significant head motion artifact. Materials and Methods: $[^{11}C]raclopride$ PET scans on 4 normal volunteers acquired with bolus plus constant infusion protocol were retrospectively analyzed. Following the 50 min resting period, the participants played a video game with a monetary reward for 40 min. Dynamic frames acquired during the equilibrium condition (pre-task: 30-50 min, task: 70-90 min, post-task: 110-120 min) were realigned to the first frame in pre-task condition. Intra-condition registrations between the frames were performed, and average image for each condition was created and registered to the pre-task image (inter-condition registration). Pre-task PET image was then co-registered to own MRI of each participant and transformation parameters were reapplied to the others. Volumes of interest (VOI) for dorsal putamen (PU) and caudate (CA), ventral striatum (VS), and cerebellum were defined on the MRI. Binding potential (BP) was measured and DAR was calculated as the percent change of BP during and after the task. SPM analyses on the BP parametric images were also performed to explore the regional difference in the effects of head motion on BP and DAR estimation. Results: Changes in position and orientation of the striatum during the PET scans were observed before the head motion correction. BP values at pre-task condition were not changed significantly after the intra-condition registration. However, the BP values during and after the task and DAR were significantly changed after the correction. SPM analysis also showed that the extent and significance of the BP differences were significantly changed by the head motion correction and such changes were prominent in periphery of the striatum. Conclusion: The results suggest that misalignment of MRI-based VOI and the striatum in PET images and incorrect DAR estimation due to the head motion during the PET activation study were significant, but could be remedied by the data-driven head motion correction.

• Korean Journal of Agricultural and Forest Meteorology
• /
• v.17 no.1
• /
• pp.35-44
• /
• 2015

Forests contain a huge amount of carbon (C) and climate change could affect forest C dynamics. This study was conducted to predict the C dynamics of Pinus densiflora and Quercus variabilis forests, which are the most dominant needleleaf and broadleaf forests in Korea, using the Korean Forest Soil Carbon (KFSC) model under the two climate change scenarios (2012-2100; Constant Temperature (CT) scenario and Representative Concentration Pathway (RCP) 8.5 scenario). To construct simulation unit, the forest land areas for those two species in the 5th National Forest Inventory (NFI) data were sorted by administrative district and stand age class. The C pools were initialized at 2012, and any disturbance was not considered during the simulation period. Although the forest C stocks of two species generally increased over time, the forest C stocks under the RCP 8.5 scenario were less than those stocks under the CT scenario. The C stocks of P. densiflora forests increased from 260.4 Tg C in 2012 to 395.3 (CT scenario) or 384.1 Tg C (RCP 8.5 scenario) in 2100. For Q. variabilis forests, the C stocks increased from 124.4 Tg C in 2012 to 219.5 (CT scenario) or 204.7 (RCP 8.5 scenario) Tg C in 2100. Compared to 5th NFI data, the initial value of C stocks in dead organic matter C pools seemed valid. Accordingly, the annual C sequestration rates of the two species over the simulation period under the RCP 8.5 scenario (65.8 and $164.2g\;C\;m^{-2}\;yr^{-1}$ for P. densiflora and Q. variabilis) were lower than those values under the CT scenario (71.1 and $193.5g\;C\;m^{-2}\;yr^{-1}$ for P. densiflora and Q. variabilis). We concluded that the C sequestration potential of P. densiflora and Q. variabilis forests could be decreased by climate change. Although there were uncertainties from parameters and model structure, this study could contribute to elucidating the C dynamics of South Korean forests in future.

• Korean Chemical Engineering Research
• /
• v.46 no.5
• /
• pp.994-1001
• /
• 2008

Freundlich isothermal adsorption parameters, applicable to such biofilter-model as process-lumping model(Lim's model), for sterilized granular activated carbon(GAC), sterilized compost and sterilized equal volume mixture of GAC and compost were obtained and were compared each other, assuming that adsorbents are enclosed by water layer, in order to construct robust process-lumping biofilter model effective for wide-range of hydrophilic volatile organic compounds(VOC). In this investigation 0.04, 0.08, 0.12, 0.16, 0.2, 0.4, 0.8 and 1.0ml of ethanol were added to three kinds of adsorbent-media and were placed at $30^{\circ}{\cdots}$ under the wet condition of the media, which was the same as biofilter operating condition, until the adsorption reached the condition of equilibrium before each adsorbed amount of ethanol was obtained. Then adsorption capacity parameters(K) and adsorption exponents of Freundlich adsorption isotherm equation, which simulates the adsorbed amount of ethanol equilibrated with the ethanol concentration of the condensed water in the pore of the media, were constructed for sterilized granular activated carbon(GAC), sterilized compost and sterilized equal volume mixture of GAC and compost as (0.7566 and $5.070{\times}10^{-7}mg-ethanol/mgmedia/(mg-ethanol/m^3)^{0.7566}$), (0.8827 and $1.000{\times}10^{-8}mg-ethanol/mgmedia/(mg-ethanol/m^3)^{0.8827}$) and (0.5688 and $5.243{\times}10^{-6}mg-ethanol/mgmedia/(mg-ethanol/m^3)^{0.5688}$), respectively. These Freundlich isothermal adsorption parameters were applicable to the adsorption characteristics of biofilter media enclosed with bio-layer. The order of magnitude of the ratio of ethanol-air/water partition coefficient and toluene-air/water partition coefficient was almost consistent to that of ethanol-adsorbed amounts in this experiment with compost and in the investigation of Delhomenie et al. on toluene-adsorption to wet compost.

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.33 no.4
• /
• pp.317-324
• /
• 2015

GNSS was firstly proposed for application in weather forecasting in the mid-1980s. It has continued to demonstrate the practical uses in GNSS meteorology, and other relevant researches are currently being conducted. Precipitable Water Vapor (PWV), calculated based on the GNSS signal delays due to the troposphere of the Earth, represents the amount of the water vapor in the atmosphere, and it is therefore widely used in the analysis of various weather phenomena such as monitoring of weather conditions and climate change detection. In this study we calculated the PWV through the meteorological information from an Automatic Weather Station (AWS) as well as GNSS data processing of a Continuously Operating Reference Station (CORS) in order to analyze the heavy snowfall of the Ulsan area in early 2014. Song’s model was adopted for the weighted mean temperature model (T_m), which is the most important parameter in the calculation of PWV. The study period is a total of 56 days (February 2013 and 2014). The average PWV of February 2014 was determined to be 11.29 mm, which is 11.34% lower than that of the heavy snowfall period. The average PWV of February 2013 was determined to be 10.34 mm, which is 8.41% lower than that of not the heavy snowfall period. In addition, certain meteorological factors obtained from AWS were compared as well, resulting in a very low correlation of 0.29 with the saturated vapor pressure calculated using the empirical formula of Magnus. The behavioral pattern of PWV has a tendency to change depending on the precipitation type, specifically, snow or rain. It was identified that the PWV showed a sudden increase and a subsequent rapid drop about 6.5 hours before precipitation. It can be concluded that the pattern analysis of GNSS PWV is an effective method to analyze the precursor phenomenon of precipitation.

• Journal of the Korean Applied Science and Technology
• /
• v.37 no.3
• /
• pp.484-495
• /
• 2020

This study is a mixed surfactant (MimicLipid-MSM1000) that forms the same structure as that of the stratum corneum, sucrose distearate, polyglyceryl-2 dioleate, fermented squalane, ergosterol, 10-hydroxystearic acid, mixture consisting of was synthesized. When using 2~5 wt% of this mixed surfactant, it was possible to make an artificial skin mimetic that forms a multi-layer lamellar structure of 5~30 layers. An emulsion was prepared using this mixed surfactant, and a multi-layered lamellar phase was formed and analyzed mechanically. The appearance of this surfactant was a light brown hard wax, the hydrophilic lipophilic balance (HLB) was 12.53, the critical parameter value was 0.987, and the acid value was 0.13. Stability according to pH change was also stable in acidic (3.8), neutral (7.2) and alkaline (10.8). The particle size of the liquid crystal was found to be the most stable maltese cross lamellar crystalline droplet at 5~25mm. The size of the emulsified particles according to the change in the speed of the homo agitator is 2500 rpm (17.9mm±2.6mm), 3500rpm (12.5mm±2.1mm), 4500rpm (6.2mm±1.8mm) particles were formed. Liquid crystal forming particles were observed through a polarization microscope, and the formation structure of the liquid crystal was precisely analyzed with a scanning electron microscope (cryo-SEM). As an application field, it is expected that it will be widely applicable to the development of various prescriptions, such as various skin care cosmetics, makeup care cosmetics, and scalp protection cosmetics, by using a skin-mimicking surfactant.