• Journal of Broadcast Engineering
• /
• v.24 no.5
• /
• pp.870-878
• /
• 2019

In this paper, we propose an Attention-based BLSTM for predicting the video compression standard of a video. Recently, in NLP, many researches have been studied to predict the next word of sentences, classify and translate sentences by their semantics using the structure of RNN, and they were commercialized as chatbots, AI speakers and translator applications, etc. LSTM is designed to solve the gradient vanishing problem in RNN, and is used in NLP. The proposed algorithm makes video compression standard prediction possible by applying BLSTM and Attention algorithm which focuses on the most important word in a sentence to a bitstream of a video, not an sentence of a natural language.

• Journal of Broadcast Engineering
• /
• v.24 no.4
• /
• pp.633-642
• /
• 2019

Recently, the convolution neural network (CNN) model at a single image super-resolution (SISR) have been very successful. The residual learning method can improve training stability and network performance in CNN. In this paper, we propose a SISR using recursive residual network architecture by introducing dense skip connections for learning nonlinear mapping from low-resolution input image to high-resolution target image. The proposed SISR method adopts a method of the recursive residual learning to mitigate the difficulty of the deep network training and remove unnecessary modules for easier to optimize in CNN layers because of the concise and compact recursive network via dense skip connection method. The proposed method not only alleviates the vanishing-gradient problem of a very deep network, but also get the outstanding performance with low complexity of neural network, which allows the neural network to perform training, thereby exhibiting improved performance of SISR method.

• Journal of Information Processing Systems
• /
• v.2 no.1
• /
• pp.52-57
• /
• 2006

One of the important functions of an Intelligent Transportation System (ITS) is to classify vehicle types using a vision system. We propose a method using machine-learning algorithms for this classification problem with 3-D object model fitting. It is also necessary to detect road lanes from a fixed traffic surveillance camera in preparation for model fitting. We apply a background mask and line analysis algorithm based on statistical measures to Hough Transform (HT) in order to remove noise and false positive road lanes. The results show that this method is quite efficient in terms of quality.

• Journal of the Optical Society of Korea
• /
• v.7 no.3
• /
• pp.174-179
• /
• 2003

We construct a four-level system where a metastable state is included in an $Er^{3＋}$ Doped Yttrium aluminum garnet (YAG) crystal. Because of the action of the coherent field, the traditional light amplification with inversion can be exhibited with remarkable variation. As a result, we propose a method to achieve the gain equalization by atomic coherence. At the same time, we find that the high index of refraction accompanied by vanishing absorption can also be reached in this model. We also find that a higher index of refraction with zero absorption can be easily obtained when the coherent field is off resonance.

• Journal of Broadcast Engineering
• /
• v.2 no.2
• /
• pp.125-135
• /
• 1997

This paper is concerned with realistic mainpulation of content in video sequences. Manipulation of content in video sequences is one of the content-based functionalities for MPEG-4 Visual standard. We present an approach to synthesizing video sequences by using the intermediate outputs of three-dimensional (3D) motion and depth analysis. For concreteness, we focus on video showing 3D motion of an observer relative to a scene containing planar runways (or roads). We first present a simple runway (or road) model. Then, we describe a method of identifying the runway (or road) boundary in the image using the Point of Heading Direction (PHD) which is defined as the image of, the ray along which a camera moves. The 3D motion of the camera is obtained from one of the existing 3D analysis methods. Then, a video sequence containing a runway is manipulated by (i) coloring the scene part above a vanishing line, say blue, to show sky, (ii) filling in the occluded scene parts, and (iii) overlaying the identified runway edges and placing yellow disks in them, simulating lights. Experimental results for a real video sequence are presented.

• Journal of information and communication convergence engineering
• /
• v.9 no.6
• /
• pp.780-784
• /
• 2011

As the popularization and development of 3D display makes common users easy to experience a solid 3D virtual reality, the demand for virtual reality contents are increasing. In this paper, we propose 3D panorama system using vanishing point locationbased depth map generation method. 3D panorama using depthmap stitching gives an effect that makes users feel staying at real place and looking around nearby circumstances. Also, 3D panorama gives free sight point for both nearby object and remote one and provides solid 3D video.

• Structural Engineering and Mechanics
• /
• v.48 no.6
• /
• pp.775-789
• /
• 2013

The current paper proposes a boundary element formulation, applicable to 2-D and 3-D elastostatics problems using a unified approach for transformations of the domain integrals into boundary integrals. The method is applicable to linear problems encompassing both finite and infinite multi-region domains allowing non-vanishing body forces. Numerical results agree quite well with the analytical solutions; while the present method offers easy formulation with less numerical efforts in comparison to FEM or some BEM which need interior points to treat arbitrary body forces. It is demonstrated that the method has the potential to have profound impact on engineering design, notably in dam-foundation interaction.

• Journal of Institute of Control, Robotics and Systems
• /
• v.21 no.4
• /
• pp.321-329
• /
• 2015

This paper addresses a new algorithm for a stereo visual odometry to measure the ego-motion of a vehicle. The new algorithm introduces an inlier grouping method based on Delaunay triangulation and vanishing point computation. Most visual odometry algorithms rely on RANSAC in choosing inliers. Those algorithms fluctuate largely in processing time between images and have different accuracy depending on the iteration number and the level of outliers. On the other hand, the new approach reduces the fluctuation in the processing time while providing accuracy corresponding to the RANSAC-based approaches.

• Annual Conference on Human and Language Technology
• /
• 2017.10a
• /
• pp.159-162
• /
• 2017

Long Short-Term Memory Recurrent Neural Network(LSTM RNN)는 순차 데이터 모델링에 적합한 딥러닝 모델이다. Bidirectional LSTM RNN(BiLSTM RNN)은 RNN의 그래디언트 소멸 문제(vanishing gradient problem)를 해결한 LSTM RNN을 입력 데이터의 양 방향에 적용시킨 것으로 입력 열의 모든 정보를 볼 수 있는 장점이 있어 자연어처리를 비롯한 다양한 분야에서 많이 사용되고 있다. Highway Network는 비선형 변환을 거치지 않은 입력 정보를 히든레이어에서 직접 사용할 수 있게 LSTM 유닛에 게이트를 추가한 딥러닝 모델이다. 본 논문에서는 Highway Network를 한국어 의미역 결정에 적용하여 기존 연구 보다 더 높은 성능을 얻을 수 있음을 보인다.

• Bulletin of the Korean Mathematical Society
• /
• v.40 no.3
• /
• pp.457-464
• /
• 2003

Given a Riemannian manifold (M, $\alpha$) with an almost Hermitian structure f and a non-vanishing covariant vector field b, consider the generalized Randers metric $L\;=\;{\alpha}+{\beta}$, where $\beta$ is a special singular Riemannian metric defined by b and f. This metric L is called an (a, b, f)-metric. We compute the inverse and the determinant of the fundamental tensor ($g_{ij}$) of an (a, b, f)-metric. Then we determine the maximal domain D of $TM{\backslash}O$ for an (a, b, f)-manifold where a y-local Finsler structure L is defined. And then we show that any (a, b, f)-manifold is quasi-C-reducible and find a condition under which an (a, b, f)-manifold is C-reducible.