• Journal of Broadcast Engineering
• /
• v.4 no.2
• /
• pp.87-102
• /
• 1999

We propose an algorithm for augmenting a real video sequence with views of graphics ojbects without metric calibration of the video camera by representing the motion of the video camera in projective space. We define a virtual camera, through which views of graphics objects are generated. attached to the real camera by specifying image locations of the world coordinate system of the virtual world. The virtual camera is decomposed into calibration and motion components in order to make full use of graphics tools. The projective motion of the real camera recovered from image matches has a function of transferring the virtual camera and makes the virtual camera move according to the motion of the real camera. The virtual camera also follows the change of the internal parameters of the real camera. This paper shows theoretical and experimental results of our application of non-metric vision to augmented reality.

• IEMEK Journal of Embedded Systems and Applications
• /
• v.19 no.1
• /
• pp.9-17
• /
• 2024

It is widely recognized that for 3D human pose estimation (HPE), dataset acquisition is expensive and the effectiveness of augmentation techniques of conventional visual recognition tasks is limited. We address these difficulties by presenting a simple but effective method that augments input images in terms of viewpoints when training a 3D human pose estimation (HPE) model. Our intuition is that meaningful variants of the input images for HPE could be obtained by viewing a human instance in the images from an arbitrary viewpoint different from that in the original images. The core idea is to synthesize new images that have self-occlusion and thus are difficult to predict at different viewpoints even with the same pose of the original example. We incorporate this idea into the training procedure of the 3D HPE model as an augmentation stage of the input samples. We show that a strategy for augmenting the synthesized example should be carefully designed in terms of the frequency of performing the augmentation and the selection of viewpoints for synthesizing the samples. To this end, we propose a new metric to measure the prediction difficulty of input images for 3D HPE in terms of the distance between corresponding keypoints on both sides of a human body. Extensive exploration of the space of augmentation probability choices and example selection according to the proposed distance metric leads to a performance gain of up to 6.2% on Human3.6M, the well-known pose estimation dataset.

• Honam Mathematical Journal
• /
• v.33 no.4
• /
• pp.557-573
• /
• 2011

Let G be a compact and connected semisimple Lie group, H a closed subgroup, g (resp. h) the Lie algebra of G (resp. H), B the Killing form of g, g the normal metric on the homogeneous space G/H which is induced by -B. Let D be an invarint connection with Weyl structure (D, g, ${\omega}$) in the tangent bundle over the normal homogeneous Riemannian manifold (G/H, g) which is projectively flat. Then, the affine connection D on (G/H, g) is a Yang-Mills connection if and only if D is the Levi-Civita connection on (G/H, g).

• ETRI Journal
• /
• v.42 no.1
• /
• pp.26-35
• /
• 2020

Herein, we consider uplink multiuser massive multiple-input multiple-output systems when multiple users transmit information symbols to a base station (BS) by applying simple space-time block coding (STBC). At the BS receiver, two detection filters for each user are used to detect the STBC information symbols. One of these filters is for odd-indexed symbols and the other for even-indexed symbols. Using constrained output variance metric minimization, we first derive a special relation between the closed-form optimal solutions for the two detection filters. Then, using the derived special relation, we propose a new blind adaptive algorithm for implementing the minimum output variance-based optimal filters. In the proposed adaptive algorithm, filter weight vectors are updated only in the region satisfying the special relation. Through a theoretical analysis of the convergence speed and a computer simulation, we demonstrate that the proposed scheme exhibits faster convergence speed and lower steady-state bit error rate than the conventional scheme.

• Journal of the Chungcheong Mathematical Society
• /
• v.13 no.1
• /
• pp.131-138
• /
• 2000

Let D be the open unit disk in the complex plane $\mathbb{C}$. For any q > 0, the operator $Q_q$ defined by $$Q_qf(z)=q\int_{D}\frac{f(\omega)}{(1-z{\bar{\omega}})^{1+q}}d{\omega},\;z{\in}D$$. maps $L^{\infty}(D)$ boundedly onto $B_q$ for each q > 0. In this paper, weighted Bloch spaces $\mathcal{B}_q$ (q > 0) are considered on the open unit ball in $\mathbb{C}^n$. In particular, we will investigate the possibility of extension of this operator to the Weighted Bloch spaces $\mathcal{B}_q$ in $\mathbb{C}^n$.

• East Asian mathematical journal
• /
• v.27 no.1
• /
• pp.1-9
• /
• 2011

In this paper, we consider an iterative scheme for finding a common element of the set of fixed points of a asymptotically quasi nonexpansive mapping and the set of solutions of the variational inequality for an inverse strongly monotone mapping in a Hilbert space. Then we show that the sequence converges strongly to a common element of two sets. Using this result, we consider the problem of finding a common fixed point of a asymptotically quasi-nonexpansive mapping and strictly pseudocontractive mapping and the problem of finding a common element of the set of fixed points of a asymptotically quasi-nonexpansive mapping and the set of zeros of an inverse-strongly monotone mapping.

• East Asian mathematical journal
• /
• v.34 no.5
• /
• pp.643-649
• /
• 2018

Let M be a connected n-dimensional submanifold of a Euclidean space $E^{n+k}$ equipped with the induced metric and ${\Delta}$ its Laplacian. If the position vector x of M is decomposed as a sum of three vectors $x=x_1+x_2+x_0$ where two vectors $x_1$ and $x_2$ are non-constant eigenvectors of the Laplacian, i.e., ${\Delta}x_i={\lambda}_ix_i$, i = 1, 2 (${\lambda}_i{\in}R$) and $x_0$ is a constant vector, then, M is called a 2-type submanifold. In this paper we proved that a connected 2-type hypersurface M in $E^{n+1}$ whose postion vector x satisfies ${\langle}{\Delta}x,x-x_0{\rangle}=c$ for a constant c, where ${\langle}$, ${\rangle}$ is the usual inner product in $E^{n+1}$, is of null 2-type and has constant mean curvature and scalar curvature.

• Journal of the Korean Mathematical Society
• /
• v.55 no.5
• /
• pp.1157-1175
• /
• 2018

For a switched system with constraint on switching sequences, which is also called a subshift action, on a metric space not necessarily compact, two kinds of topological entropies, average topological entropy and maximal topological entropy, are introduced. Then we give some properties of those topological entropies and estimate the bounds of them for some special systems, such as subshift actions generated by finite smooth maps on p-dimensional Riemannian manifold and by a family of surjective endomorphisms on a compact metrizable group. In particular, for linear switched systems on ${\mathbb{R}}^p$, we obtain a better upper bound, by joint spectral radius, which is sharper than that by Wang et al. in [42,43].

• Journal of Korea Multimedia Society
• /
• v.23 no.8
• /
• pp.927-939
• /
• 2020

Gait recognition can identify people's identity from a long distance, which is very important for improving the intelligence of the monitoring system. Among many human features, gait features have the advantages of being remotely available, robust, and secure. Traditional gait feature extraction, affected by the development of behavior recognition, can only rely on manual feature extraction, which cannot meet the needs of fine gait recognition. The emergence of deep convolutional neural networks has made researchers get rid of complex feature design engineering, and can automatically learn available features through data, which has been widely used. In this paper,conduct feature metric learning in the three-dimensional space by combining the three-dimensional convolution features of the gait sequence and the Siamese structure. This method can capture the information of spatial dimension and time dimension from the continuous periodic gait sequence, and further improve the accuracy and practicability of gait recognition.

• ETRI Journal
• /
• v.30 no.5
• /
• pp.663-673
• /
• 2008

The multi-scenario topology of multi-channel hybrid ubiquitous sensor networks (USNs) is studied and a novel link auto-diversity cross-layer cooperative scheduling scheme is proposed in this paper. The proposed scheme integrates the attributes of the new performance evaluation link auto-diversity air-time metric and the topology space in the given multi-scenario. The proposed scheme is compared with other schemes, and its superiority is demonstrated through simulations. The simulation results show that relative energy consumption, link reception probability, and end-to-end blocking probability are improved. The addressing ratio of success with unchanged parameters and external information can be increased. The network can tolerate more hops to support reliable transportation when the proposed scheme is implemented. Moreover, the scheme can make the network stable. Therefore, the proposed scheme can enhance the average rate performance of the hybrid USN and stabilize the outage probability.