Abstract: Human pose estimation and tracking are to accurately identify and locate the positions of human joints in the video. It is a computer vision task which is of great significance for human motion recognition, behavior understanding and scene analysis. There has been remarkable progress on human pose estimation in recent years. However, more researches are needed for human pose tracking especially for online tracking. In this paper, a framework, called PoseSRPN, is proposed for online single-person pose estimation and tracking. We use Siamese network attaching a pose estimation branch to incorporate Single-person Pose Tracking (SPT) and Visual Object Tracking (VOT) into one framework. The pose estimation branch has a simple network structure that replaces the complex upsampling and convolution network structure with deconvolution. By augmenting the loss of fully convolutional Siamese network with the pose estimation task, pose estimation and tracking can be trained in one stage. Once trained, PoseSRPN only relies on a single bounding box initialization and producing human joints location. The experimental results show that while maintaining the good accuracy of pose estimation on COCO and PoseTrack datasets, the proposed method achieves a speed of 59 frame/s, which is superior to other pose tracking frameworks.
Abstract: In seismic data processing, attenuation of random noise
is the basic step to improve quality of data for further application
of seismic data in exploration and development in different gas
and oil industries. The signal-to-noise ratio of the data also highly
determines quality of seismic data. This factor affects the reliability
as well as the accuracy of seismic signal during interpretation
for different purposes in different companies. To use seismic data
for further application and interpretation, we need to improve the
signal-to-noise ration while attenuating random noise effectively.
To improve the signal-to-noise ration and attenuating seismic
random noise by preserving important features and information
about seismic signals, we introduce the concept of anisotropic
total fractional order denoising algorithm. The anisotropic total
fractional order variation model defined in fractional order bounded
variation is proposed as a regularization in seismic denoising. The
split Bregman algorithm is employed to solve the minimization
problem of the anisotropic total fractional order variation model
and the corresponding denoising algorithm for the proposed method
is derived. We test the effectiveness of theproposed method for
synthetic and real seismic data sets and the denoised result is
compared with F-X deconvolution and non-local means denoising
algorithm.
Abstract: NiFe2O4 (nickel ferrite), ZnFe2O4 (zinc ferrite) and
Ni0.5Zn0.5Fe2O4 (nickel-zinc ferrite) were prepared by
mechanochemical route in a planetary ball mill starting from mixture
of the appropriate quantities of the Ni(OH)2/Fe(OH)3,
Zn(OH)2/Fe(OH)3 and Ni(OH)2/Zn(OH)2/Fe(OH)3 hydroxide
powders. In order to monitor the progress of chemical reaction and
confirm phase formation, powder samples obtained after 25 h, 18 h
and 10 h of milling were characterized by X-ray diffraction (XRD),
transmission electron microscopy (TEM), IR, Raman and Mössbauer
spectroscopy. It is shown that the soft mechanochemical method, i.e.
mechanochemical activation of hydroxides, produces high quality
single phase ferrite samples in much more efficient way. From the IR
spectroscopy of single phase samples it is obvious that energy of
modes depends on the ratio of cations. It is obvious that all samples
have more than 5 Raman active modes predicted by group theory in
the normal spinel structure. Deconvolution of measured spectra
allows one to conclude that all complex bands in the spectra are made
of individual peaks with the intensities that vary from spectrum to
spectrum. The deconvolution of Raman spectra allows to separate
contributions of different cations to a particular type of vibration and
to estimate the degree of inversion.
Abstract: In this paper, the improvement by deconvolution of
the depth resolution in Secondary Ion Mass Spectrometry (SIMS)
analysis is considered. Indeed, we have developed a new Tikhonov-
Miller deconvolution algorithm where a priori model of the solution
is included. This is a denoisy and pre-deconvoluted signal obtained
from: firstly, by the application of wavelet shrinkage algorithm,
secondly by the introduction of the obtained denoisy signal in an
iterative deconvolution algorithm. In particular, we have focused the
light on the effect of the iterations number on the evolution of the
deconvoluted signals. The SIMS profiles are multilayers of Boron in
Silicon matrix.
Abstract: We present a numerical study of the sensitivity of the so called time relaxation family of models of fluid motion with respect to the time relaxation parameter χ on the two dimensional cavity problem. The goal of the study is to compute and compare the sensitivity of the model using finite difference method (FFD) and sensitivity equation method (SEM).
Abstract: The increments of aromatic structures are widely used to monitor the degree of humification. Compost derived from mix manures mixed with agricultural wastes was studied. The compost collected at day 0, 7, 14, 21, 28, 35, 49, 77, 91, 105, and 119 was divided into 3 stages, initial stage at day 0, thermophilic stage during day 1-48, and mature stage during day 49-119. The change of highest absorptions at wavelength range between 210-235 nm during day 0- 49 implied that small molecules such as nitrates and carboxylic occurred faster than the aromatic molecules that were found at wavelength around 280 nm. The ratio of electron-transfer band at wavelength 253 nm by the benzonoid band at wavelength 230 nm (E253/E230) also gradually increased during the fermenting period indicating the presence of O-containing functional groups. This was in agreement with the shift change from aliphatic to aromatic structures as shown by the relationship with C/N and H/C ratios (r = - 0.631 and -0.717, p< 0.05) since both were decreasing. Although the amounts of humic acid (HA) were not different much during the humification process, the UV spectral deconvolution showed better qualitative characteristics to help in determining the compost quality. From this study, the compost should be used at day 49 and should not be kept longer than 3 months otherwise the quality of HA would decline regardless of the amounts of HA that might be rising. This implied that other processes, such as mineralization had an influence on the humification process changing HA-s structure and its qualities.
Abstract: This paper presents the doping profile measurement
and characterization technique for the pocket implanted nano scale
n-MOSFET. Scanning capacitance microscopy and atomic force
microscopy have been used to image the extent of lateral dopant
diffusion in MOS structures. The data are capacitance vs. voltage
measurements made on a nano scale device. The technique is nondestructive
when imaging uncleaved samples. Experimental data from
the published literature are presented here on actual, cleaved device
structures which clearly indicate the two-dimensional dopant profile
in terms of a spatially varying modulated capacitance signal. Firstorder
deconvolution indicates the technique has much promise for
the quantitative characterization of lateral dopant profiles. The pocket
profile is modeled assuming the linear pocket profiles at the source
and drain edges. From the model, the effective doping concentration
is found to use in modeling and simulation results of the various
parameters of the pocket implanted nano scale n-MOSFET. The
potential of the technique to characterize important device related
phenomena on a local scale is also discussed.
Abstract: This paper presents a generalized form of the
mechanistic deconvolution technique (GMD) to modeling image sensors applicable in various pan–tilt planes of view. The mechanistic deconvolution technique (UMD) is modified with the
given angles of a pan–tilt plane of view to formulate constraint parameters and characterize distortion effects, and thereby, determine
the corrected image data. This, as a result, does not require experimental setup or calibration. Due to the mechanistic nature of
the sensor model, the necessity for the sensor image plane to be
orthogonal to its z-axis is eliminated, and it reduces the dependency on image data. An experiment was constructed to evaluate the
accuracy of a model created by GMD and its insensitivity to changes in sensor properties and in pan and tilt angles. This was compared
with a pre-calibrated model and a model created by UMD using two sensors with different specifications. It achieved similar accuracy
with one-seventh the number of iterations and attained lower mean error by a factor of 2.4 when compared to the pre-calibrated and
UMD model respectively. The model has also shown itself to be robust and, in comparison to pre-calibrated and UMD model, improved the accuracy significantly.
Abstract: This work explores blind image deconvolution by recursive function approximation based on supervised learning of neural networks, under the assumption that a degraded image is linear convolution of an original source image through a linear shift-invariant (LSI) blurring matrix. Supervised learning of neural networks of radial basis functions (RBF) is employed to construct an embedded recursive function within a blurring image, try to extract non-deterministic component of an original source image, and use them to estimate hyper parameters of a linear image degradation model. Based on the estimated blurring matrix, reconstruction of an original source image from a blurred image is further resolved by an annealed Hopfield neural network. By numerical simulations, the proposed novel method is shown effective for faithful estimation of an unknown blurring matrix and restoration of an original source image.
Abstract: The characterisation of agro-wastes fibres for composite applications from Nigeria using X-ray diffraction (XRD) and Scanning Electron Microscopy (SEM) has been done. Fibres extracted from groundnut shell, coconut husk, rice husk, palm fruit bunch and palm fruit stalk are processed using two novel cellulose fibre production methods developed by the authors. Cellulose apparent crystallinity calculated using the deconvolution of the diffractometer trace shows that the amorphous portion of cellulose was permeable to hydrolysis yielding high crystallinity after treatment. All diffratograms show typical cellulose structure with well-defined 110, 200 and 040 peaks. Palm fruit fibres had the highest 200 crystalline cellulose peaks compared to others and it is an indication of rich cellulose content. Surface examination of the resulting fibres using SEM indicates the presence of regular cellulose network structure with some agglomerated laminated layer of thin leaves of cellulose microfibrils. The surfaces were relatively smooth indicating the removal of hemicellulose, lignin and pectin.
Abstract: A frequency grouping approach for multi-channel
instantaneous blind source separation (I-BSS) of convolutive
mixtures is proposed for a lower net residual inter-symbol
interference (ISI) and inter-channel interference (ICI) than the
conventional short-time Fourier transform (STFT) approach. Starting
in the time domain, STFTs are taken with overlapping windows to
convert the convolutive mixing problem into frequency domain
instantaneous mixing. Mixture samples at the same frequency but
from different STFT windows are grouped together forming unique
frequency groups.
The individual frequency group vectors are input to the I-BSS
algorithm of choice, from which the output samples are dispersed
back to their respective STFT windows. After applying the inverse
STFT, the resulting time domain signals are used to construct the
complete source estimates via the weighted overlap-add method
(WOLA). The proposed algorithm is tested for source deconvolution
given two mixtures, and simulated along with the STFT approach to
illustrate its superiority for fairly motionless sources.