• Journal of Internet of Things and Convergence
• /
• v.9 no.6
• /
• pp.111-117
• /
• 2023

There is a lot of research on using a combination of graph neural networks and recurrent neural networks as a way to account for both temporal and spatial dependencies. In particular, graph neural networks are an emerging area of research. Seoul's bicycle rental service (aka Daereungi) has rental stations all over the city of Seoul, and the rental information at each station is a time series that is faithfully recorded. The rental information of each rental station has temporal characteristics that show periodicity over time, and regional characteristics are also thought to have important effects on the rental status. Regional correlations can be well understood using graph neural networks. In this study, we reconstructed the time series data of Seoul's bicycle rental service into a graph and developed a rental prediction model that combines a graph neural network and a recurrent neural network. We considered temporal characteristics such as periodicity over time, regional characteristics, and the degree importance of each rental station.

• ALGAE
• /
• v.35 no.3
• /
• pp.225-236
• /
• 2020

Gymnodinium smaydae is a newly described mixotrophic dinoflagellate that feeds on only Heterocapsa spp. and Scrippsiella acuminata among 19 tested algal prey. It is one of the fastest growing dinoflagellates when feeding, but does not grow well without prey. To investigate its spatial-temporal distributions in Korean waters, we quantified its abundance in water samples that were seasonally collected from 28 stations along the Korean Peninsula from April 2015 to October 2018, using quantitative real-time polymerase chain reactions. This dinoflagellate had a wide distribution, as reflected by the detection of G. smaydae cells at 23 of the sampling stations. However, this distribution had a strong seasonality; it was detected at 21 stations in the summer and only one station in winter. The abundance of G. smaydae was significantly and positively correlated with chlorophyll a concentration as well as with water temperature. However, there were no significant correlations between the abundance of G. smaydae and salinity, concentrations of nutrients, or dissolved oxygen concentration. During the study period, G. smaydae was present when water temperatures were 7.6-28.0℃, salinities were 9.6-34.1, concentrations of NO₃ were not detectable-106.0 μM, and concentrations of PO₄ were not detectable-3.4 μM. The highest abundance of G. smaydae was 18.5 cells mL^-1 in the coastal waters of Jinhae in July 2017 when the chlorophyll a concentration was 127 mg m^-3 and water temperature was 23.8℃. Therefore, the spatial-temporal distribution of G. smaydae in Korean coastal waters may be affected by chlorophyll a concentration and water temperature.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.25 no.3B
• /
• pp.408-422
• /
• 2000

In this paper, we propose a generalized vector channel model for wireless communication systems using antenna arrays. The proposed channel model reflects path loss. spatial-temporal variation of shadowing, multipath fading , Doppler effect, spatial distribution of local scatterers and delay spread due to remote dominant scatterers. In addition, we use a discrete ray model in which respective ray signal experiences independent shadowing, fading and Doppler shift, and impinges on antenna arrays at a distinct angle. Based on the proposed mode. we derive the relations on the spatial and temporal correlations of the received signals and implement a spatial channel simulator. By comparing the theoretical values with the simulated ones, we verify the effectiveness of the implemented simulator. The simulator is then used to generate an arbitrary channel impulse response and to analyze the channel characteristics under various environments.

• Physical Therapy Korea
• /
• v.22 no.2
• /
• pp.1-10
• /
• 2015

The common features of walking in patients with stroke include decreased gait velocity and increased asymmetrical gait pattern. The purpose of this study was to identify important factors related to impairments in gait velocity and asymmetry in chronic stroke patients. The subjects were 30 independently ambulating subjects with chronic stroke. The subjects' impairments were examined, including the isokinetic peak torque of knee extensors, knee flexors, ankle plantarflexors, and ankle dorsiflexors. Passive and active ranges of motion (ROM) of the ankle joint, ankle plantarflexor spasticity, joint position senses of the knee and ankle joint, and balance were examined together. In addition, gait velocity and temporal and spatial asymmetry were evaluated with subjects walking at their comfortable speed. Pearson correlations and multiple regressions were used to measure the relationships between impairments and gait speed and impairments and asymmetry. Regression analyses revealed that ankle passive ROM and peak torque of knee flexors were important factors for gait velocity ($R^2=.41$), while ankle passive ROM was the most important determinant for temporal asymmetry ($R^2=.35$). In addition, knee extensor peak torque was the most significant factor for gait spatial asymmetry ($R^2=.17$). Limitation in ankle passive ROM and weakness of the knee flexor were major contributors to slow gait velocity. Moreover, limited passive ROM in the ankle influenced the level of temporal gait asymmetry in chronic stroke patients. Our findings suggest that stroke rehabilitation programs aiming to improve gait velocity and temporal asymmetry should include stretching exercise for the ankle joint.

• The Journal of the Korea Contents Association
• /
• v.18 no.12
• /
• pp.35-45
• /
• 2018

Motion Estimation(ME) has been developed to reduce the redundant data in digital video signal. ME is an important part of video encoding system, However, it requires huge computational complexity of the encoder part, and fast motion search methods have been proposed to reduce huge complexity. Multi- view video is obtained by capturing on a three-dimensional scene with many cameras at different positions and its complexity increases in proportion to the number of cameras. In this paper, we proposed an efficient motion method which chooses a search pattern adaptively by using the temporal-spatial correlation of the block and the characteristics of the block. Experiment results show that the computational complexity reduction of the proposed method over TZ search method and FS method can be up to 70~75% and 99% respectively while keeping similar image quality and bit rates.

• Journal of KIISE:Software and Applications
• /
• v.31 no.9
• /
• pp.1142-1149
• /
• 2004

Motion Estimation(ME) has been developed to reduce temporal redundancy in digital video signals and increase data compression ratio. ME is an Important part of video encoding systems, since it can significantly affect the output quality of encoded sequences. However, ME requires high computational complexity, it is difficult to apply to real time video transmission. for this reason, motion estimation algorithms with low computational complexity are viable solutions. In this paper, we present an efficient method with low computational complexity based on spatial and temporal correlations of motion vectors. The proposed method uses temporally and spatially correlated motion information, the motion vector of the block with the same coordinate in the reference frame and the motion vectors of neighboring blocks around the current block in the current frame, to decide the search pattern and the location of search starting point adaptively. Experiments show that the image quality improvement of the proposed method over MVFAST (Motion Vector Field Adaptive Search Technique) and PMVFAST (Predictive Motion Vector Field Adaptive Search Technique) is 0.01～0.3(dB) better and the speedup improvement is about 1.12～l.33 times faster which resulted from lower computational complexity.

• Korean Journal of Fisheries and Aquatic Sciences
• /
• v.39 no.spc1
• /
• pp.214-222
• /
• 2006

Dominant species of macrobenthos were analyzed based on differentiation of three distinct methods: the density based method, the biomass based method and LeBris method, by considering the frequency of occurrence using quantitative data collected over 5 years (1993-1998) at 21 stations in Chonsu Bay. Sedimentary environments as well as species composition and diversity showed the spatial and temporal distribution patterns. The ranks of dominant species as determined by the density based method were more similar to the results by the LeBris method than to those from the biomass based method. Considering the temporal variation, LeBris method were more efficient than any other methods for the determination of dominant species in Chonsu Bay. Lumbrineris longifolia, Theora fragilis, and Moerella jedoensis were recognized by all three methods. A one-way analysis of variance indicated spatial distributions patterns among most of the dominant species. These species showed positive correlations to sedimentary parameters such as mean grain size. However, T. fragilis and Paraprinospio pinnata showed the temporal patterns in their distribution, and were also correlated to the benthic environment, organic content and dissolved oxygen. Some dominant species, e.g., T. fragilis, S. scutata, G. gurjanovae proved to be useful benthic indicators based on the environmental variations determinated by long-term benthic ecological monitoring in Chonsu Bay.

• Proceedings of the Korean Society of Soil and Groundwater Environment Conference
• /
• 2003.04a
• /
• pp.63-66
• /
• 2003

This study applied a hydrogeological field survey and isotope investigation to identify source locations and delineate pathways of groundwater contamination by nitrogen compounds. The infiltration and recharge processes were analyzed with groundwater-level fluctuation data and oxygen-hydrogen stable isotope data. The groundwater flow pattern was investigated through groundwater flow modeling and spatial and temporal variation of oxygen isotope data. Based on the flow analysis and nitrogen isotope data, source types of nitrate contamination in groundwater are identified. Groundwater recharge largely occurs in spring and summer due to precipitation or irrigation water in rice fields. Based on oxygen isotope data and cross-correlation between precipitation and groundwater level changes, groundwater recharge was found to be mainly caused by irrigation in spring and by precipitation at other times. The groundwater flow velocity calculated by a time series of spatial correlations, 231 m/yr, is in good accordance with the linear velocity estimated from hydrogeologic data. Nitrate contamination sources are natural and fertilized soils as non-point sources, and septic and animal wastes as point sources. Seasonal loading and spatial distribution of nitrate sources are estimated by using oxygen and nitrogen isotopic data.

• The Journal of the Korea Contents Association
• /
• v.12 no.4
• /
• pp.31-39
• /
• 2012

DVC(Distributed Video Coding) techniques have been attracting a lot of research works since these enable us to implement the light-weight video encoder and to provide good coding efficiency by introducing the feedback channel. However, the feedback channel causes the decoder to increase the decoding complexity and requires very high decoding latency because of numerous iterative decoding processes. So, in order to reduce the decoding delay and then to implement in a real-time environment, this paper proposes several parity bit estimation methods which are based on the temporal correlation, spatial correlation and spatio-temporal correlations between LDPCA frames on each bit plane in the consecutive video frames in pixel-domain Wyner-Ziv video coding scheme and then the performances of these methods are compared in fast DVC scheme. Through computer simulations, it is shown that the adaptive spatio-temporal correlation-based estimation method and the temporal correlation-based estimation method outperform others for the video frames with the highly active contents and the low active contents, respectively. By using these results, the proposed estimation schemes will be able to be effectively used in a variety of different applications.

• Wind and Structures
• /
• v.4 no.5
• /
• pp.413-440
• /
• 2001

This paper presents an extensive analysis of unsteady wind loading data on a 18 m long and 2 m high wall in a rural environment, with the wind at a range of angles to the wall normal. The data is firstly analyzed using standard statistical techniques (moments of probability distributions, auto- and cross-correlations, auto- and cross-spectra etc.). The analysis is taken further using a variety of less conventional methods - conditional sampling, proper orthogonal decomposition and wavelet analysis. It is shown that, even though the geometry is simple, the nature of the unsteady flow is surprisingly complex. The fluctuating pressures on the front face of the wall are to a great extent caused by the turbulent fluctuations in the upstream flow, and reflect the oncoming flow structures. The results further suggest that there are distinct structures in the oncoming flow with a variety of scales, and that the second order quasi-steady approach can predict the pressure fluctuations quite well. The fluctuating pressures on the rear face are also influenced by the fluctuations in the oncoming turbulence, but also by unsteady fluctuations due to wake unsteadiness. These fluctuations have a greater temporal and spatial coherence than on the front face and the quasi-steady method over-predicts the extent of these fluctuations. Finally the results are used to check some assumptions made in the current UK wind loading code of practice.