To measure the influence of depth estimation error on depth coding and virtual view rendering in a free-viewpoint television (FTV) system, a multi-view depth estimation is proposed to solve the problem for coding and ...
详细信息
To measure the influence of depth estimation error on depth coding and virtual view rendering in a free-viewpoint television (FTV) system, a multi-view depth estimation is proposed to solve the problem for coding and view synthesis. Firstly, the consistency of initial depth is checked and the influence of initial miss-matches is minimized by introducing an additional adaptive matching error that penalizes the unreliable matches. Then according to specific criteria, the multi-reference depth maps are merged into one disparity map to improve the quality of disparity map. Finally, a multi-lateral filtering is used to preserve details in the depth map and simultaneously smooth the depths in occluded areas at object boundary, less texture and discontinuity regions. Experimental results show a significant improvement of the initial input depth maps and coding efficiency, as well as a reduction of view synthesis artifacts.
In this research, we use S-system model and Differential Evolution based inference method to capture cellular dynamics using available mutual interaction information among genes. We propose a new fitness function, eff...
详细信息
ISBN:
(纸本)9781605587226
In this research, we use S-system model and Differential Evolution based inference method to capture cellular dynamics using available mutual interaction information among genes. We propose a new fitness function, effectively incorporating a priori information, which guides the inference method to deduce correct skeletal structure of the network with more accurate parameter values. Proposed fitness function mirrors user's confidence in the validity of knowledge and helps in narrowing down the search range of the model parameters for highly confident knowledge. We investigate the potency of the method in terms of quality of data and required data size. The proposed method is shown to perform better in inherent noisy data and in presence of small number of time-dynamics data. We also investigate how the inference method performs in terms of iterative incorporation of knowledge. In inferring cell-cycle data of budding yeast (Saccharomyces cerevisiae), guided by knowledge, the inference method predicts 17 and 23 correct regulations in first and second iteration, respectively which is significantly higher than some other existing methods. Along with finding the parameter values more accurately, it predicts some new regulations and helps in revealing the underlying network structure. Copyright 2010 ACM.
The time for new paradigms for pay-per-use road use charging (RUC) has come, and it is still unclear whether or not current Global Navigation Satellite systems (GNSS) supply reliability enough to lead the technologica...
详细信息
Orthogonal moment functions have long been used in image analysis. This paper proposes a novel approach based on discrete orthogonal Tchebichef Moment for efficient image compression. The method incorporates simplifie...
详细信息
Medical images are generally of poor contrast and complex types of noise and blur. So denoising of images is an important role in processing and analysis of medical image. The purpose of the paper takes advantages of ...
详细信息
Passwords are no longer valid for accessing personal resources, because it is easy to forge them, in either small or big systems. Today's challenge is to find other alternatives for passwords. Biometric is one of ...
详细信息
In symbiotic computing approach, digital world and humans interact with each other, based on "basic principle of symbiosis", where they increase information processing ability, activity, and stability by off...
详细信息
The exposure of human body to electric field and magnetic field could cause biological effects, including changes in functions of cells and tissues and subtle changes in hormone levels, which may or may not be harmful...
详细信息
The significance of the distant learning is not a novelty, yet in terms of the study effectiveness the conventional e-learning is not sufficient. A new model of learning is based on the new paradigm -the personalizati...
详细信息
ISBN:
(纸本)9781936338054
The significance of the distant learning is not a novelty, yet in terms of the study effectiveness the conventional e-learning is not sufficient. A new model of learning is based on the new paradigm -the personalization of teaching environment taking into consideration the personal parameters of students, their abilities, learning styles, etc. That possibility is offered by current systems only in a very limited scope. The paper deals with the description of an individualized e-learning system which automatically adapts the different characteristics and learning styles of students. The system consists of three modules: Within the Student Module, the set of mutually independent student characteristics influencing the process of their learning styles has been designed on the basis of analysis of published classifications. For the Virtual Teacher Module variants of teaching procedures wieldy via e-learning form corresponding with the determined student characteristics have been designed. And finally, the Adaptive Module is represented by a set of adaptive algorithms selecting optimal study materials for individual types of students. All those mechanisms will be applied in the new adaptive LMS. The contribution involves a pilot description of the proposed structure, including the description of learning styles analysis and its results, and finally, next gradual development of the issue being solved is formulated here.
暂无评论