Abstract: We propose to record Activities of Daily Living
(ADLs) of elderly people using a vision-based system so as to provide
better assistive and personalization technologies. Current ADL-related
research is based on data collected with help from non-elderly subjects
in laboratory environments and the activities performed are predetermined
for the sole purpose of data collection. To obtain more
realistic datasets for the application, we recorded ADLs for the elderly
with data collected from real-world environment involving real elderly
subjects. Motivated by the need to collect data for more effective
research related to elderly care, we chose to collect data in the room of
an elderly person. Specifically, we installed Kinect, a vision-based
sensor on the ceiling, to capture the activities that the elderly subject
performs in the morning every day. Based on the data, we identified
12 morning activities that the elderly person performs daily. To
recognize these activities, we created a HARELCARE framework to
investigate into the effectiveness of existing Human Activity
Recognition (HAR) algorithms and propose the use of a transfer
learning algorithm for HAR. We compared the performance, in terms
of accuracy, and training progress. Although the collected dataset is
relatively small, the proposed algorithm has a good potential to be
applied to all daily routine activities for healthcare purposes such as
evidence-based diagnosis and treatment.
Abstract: Big Data has been attracted a lot of attentions in many fields for analyzing research issues based on a large number of maternal data. Electronic Toll Collection (ETC) is one of Intelligent Transportation System (ITS) applications in Taiwan, used to record starting point, end point, distance and travel time of vehicle on the national freeway. This study, taking advantage of ETC big data, combined with urban planning theory, attempts to explore various phenomena of inter-city transportation activities. ETC, one of government's open data, is numerous, complete and quick-update. One may recall that living area has been delimited with location, population, area and subjective consciousness. However, these factors cannot appropriately reflect what people’s movement path is in daily life. In this study, the concept of "Living Area" is replaced by "Influence Range" to show dynamic and variation with time and purposes of activities. This study uses data mining with Python and Excel, and visualizes the number of trips with GIS to explore influence range of Tainan city and the purpose of trips, and discuss living area delimited in current. It dialogues between the concepts of "Central Place Theory" and "Living Area", presents the new point of view, integrates the application of big data, urban planning and transportation. The finding will be valuable for resource allocation and land apportionment of spatial planning.
Abstract: It is an indispensible strategy to adopt greenery
approach on architectural bases so as to improve ecological habitats,
decrease heat-island effect, purify air quality, and relieve surface
runoff as well as noise pollution, all of which are done in an attempt to
achieve sustainable environment. How we can do with plant design to
attain the best visual quality and ideal carbon dioxide fixation depends
on whether or not we can appropriately make use of greenery
according to the nature of architectural bases. To achieve the goal, it is
a need that architects and landscape architects should be provided with
sufficient local references. Current greenery studies focus mainly on
the heat-island effect of urban with large scale. Most of the architects
still rely on people with years of expertise regarding the adoption and
disposition of plantation in connection with microclimate scale.
Therefore, environmental design, which integrates science and
aesthetics, requires fundamental research on landscape environment
technology divided from building environment technology. By doing
so, we can create mutual benefits between green building and the
environment. This issue is extremely important for the greening design
of the bases of green buildings in cities and various open spaces. The
purpose of this study is to establish plant selection and allocation
strategies under different building sunshade levels. Initially, with the
shading of sunshine on the greening bases as the starting point, the
effects of the shades produced by different building types on the
greening strategies were analyzed. Then, by measuring the PAR
(photosynthetic active radiation), the relative DLI (daily light integral)
was calculated, while the DLI Map was established in order to
evaluate the effects of the building shading on the established
environmental greening, thereby serving as a reference for plant
selection and allocation. The discussion results were to be applied in
the evaluation of environment greening of greening buildings and
establish the “right plant, right place” design strategy of multi-level
ecological greening for application in urban design and landscape
design development, as well as the greening criteria to feedback to the
eco-city greening buildings.
Abstract: The novel 3D SnO cabbages self-assembled by
nanosheets were successfully synthesized via template-free
hydrothermal growth method under facile conditions. The XRD
results manifest that the as-prepared SnO is tetragonal phase. The
TEM and HRTEM results show that the cabbage nanosheets are
polycrystalline structure consisted of considerable single-crystalline
nanoparticles. Two typical Raman modes A1g=210 and Eg=112 cm-1
of SnO are observed by Raman spectroscopy. Moreover, galvanostatic
cycling tests has been performed using the SnO cabbages as anode
material of lithium ion battery and the electrochemical results suggest
that the synthesized SnO cabbage structures are a promising anode
material for lithium ion batteries.
Abstract: This paper presents an optimization method for
reducing the number of input channels and the complexity of the
feed-forward NARX neural network (NN) without compromising the
accuracy of the NN model. By utilizing the correlation analysis
method, the most significant regressors are selected to form the input
layer of the NN structure. An application of vehicle dynamic model
identification is also presented in this paper to demonstrate the
optimization technique and the optimal input layer structure and the
optimal number of neurons for the neural network is investigated.
Abstract: Verapamil has been shown to inhibit fentanyl uptake in vitro and is a potent P-glycoprotein inhibitor. Tissue partitioning of loperamide, a commercially available opioid, is closely controlled by the P-gp efflux transporter. The following studies were designed to evaluate the effect of opioids on verapamil partitioning in the lung and brain, in vivo. Opioid (fentanyl or loperamide) was administered by intravenous infusion to Sprague Dawley rats alone or in combination with verapamil and plasma, with lung and brain tissues were collected at 1, 5, 6, 8, 10 and 60 minutes. Drug dispositions were modeled by recirculatory pharmacokinetic models. Fentanyl slightly increased the verapamil lung (PL) partition coefficient yet decreased the brain (PB) partition coefficient. Furthermore, loperamide significantly increased PLand PB. Fentanyl reduced the verapamil volume of distribution (V1) and verapamil elimination clearance (ClE). Fentanyl decreased verapamil brain partitioning, yet increased verapamil lung partitioning. Also, loperamide increased lung and brain partitioning in vivo. These results suggest that verapamil and fentanyl may be substrates of an unidentified inward transporter in brain tissue and confirm that verapamil and loperamide are substrates of the efflux transporter P-gp.
Abstract: In this paper, we have proposed a novel FinFET with
extended body under the poly gate, which is called EB-FinFET, and
its characteristic is demonstrated by using three-dimensional (3-D)
numerical simulation. We have analyzed and compared it with
conventional FinFET. The extended body height dependence on the
drain induced barrier lowering (DIBL) and subthreshold swing (S.S)
have been also investigated. According to the 3-D numerical
simulation, the proposed structure has a firm structure, an acceptable
short channel effect (SCE), a reduced series resistance, an increased
on state drain current (I
on) and a large normalized I
DS. Furthermore,
the structure can also improve corner effect and reduce self-heating
effect due to the extended body. Our results show that the EBFinFET
is excellent for nanoscale device.
Abstract: The MFCAV Riemann solver is practically used in many Lagrangian or ALE methods due to its merit of sharp shock profiles and rarefaction corners, though very often with numerical oscillations. By viewing it as a modification of the WWAM Riemann solver, we apply the MFCAV Riemann solver to the Lagrangian method recently developed by Maire. P. H et. al.. The numerical experiments show that the application is successful in that the shock profiles and rarefaction corners are sharpened compared with results obtained using other Riemann solvers. Though there are still numerical oscillations, they are within the range of the MFCAV applied in onther Lagrangian methods.
Abstract: In this paper, we propose a novel metal oxide
semiconductor field effect transistor with L-shaped channel structure
(LMOS), and several type of L-shaped structures are also designed,
studied and compared with the conventional MOSFET device for the
same average gate length (Lavg). The proposed device electrical
characteristics are analyzed and evaluated by three dimension (3-D)
ISE-TCAD simulator. It can be confirmed that the LMOS devices
have higher on-state drain current and both lower drain-induced
barrier lowering (DIBL) and subthreshold swing (S.S.) than its
conventional counterpart has. In addition, the transconductance and
voltage gain properties of the LMOS are also improved.
Abstract: This study is to investigate the electroencephalogram (EEG) differences generated from a normal and Alzheimer-s disease (AD) sources. We also investigate the effects of brain tissue distortions due to AD on EEG. We develop a realistic head model from T1 weighted magnetic resonance imaging (MRI) using finite element method (FEM) for normal source (somatosensory cortex (SC) in parietal lobe) and AD sources (right amygdala (RA) and left amygdala (LA) in medial temporal lobe). Then, we compare the AD sourced EEGs to the SC sourced EEG for studying the nature of potential changes due to sources and 5% to 20% brain tissue distortions. We find an average of 0.15 magnification errors produced by AD sourced EEGs. Different brain tissue distortion models also generate the maximum 0.07 magnification. EEGs obtained from AD sources and different brain tissue distortion levels vary scalp potentials from normal source, and the electrodes residing in parietal and temporal lobes are more sensitive than other electrodes for AD sourced EEG.
Abstract: Aerial and satellite images are information rich. They are also complex to analyze. For GIS systems, many features require fast and reliable extraction of roads and intersections. In this paper, we study efficient and reliable automatic extraction algorithms to address some difficult issues that are commonly seen in high resolution aerial and satellite images, nonetheless not well addressed in existing solutions, such as blurring, broken or missing road boundaries, lack of road profiles, heavy shadows, and interfering surrounding objects. The new scheme is based on a new method, namely reference circle, to properly identify the pixels that belong to the same road and use this information to recover the whole road network. This feature is invariable to the shape and direction of roads and tolerates heavy noise and disturbances. Road extraction based on reference circles is much more noise tolerant and flexible than the previous edge-detection based algorithms. The scheme is able to extract roads reliably from images with complex contents and heavy obstructions, such as the high resolution aerial/satellite images available from Google maps.