A Novel Rule Refinement Method for SMT through Simulated Post-Editing

Author(s):  
Sitong Yang ◽  
Heng Yu ◽  
Qun Liu
2015 ◽  
Vol 61 ◽  
pp. 5-22
Author(s):  
Sir Dai Rees

Struther Arnott worked tirelessly as a researcher, teacher, leader and maker and implementer of policy in universities in Britain and the USA, always carrying his colleagues along with him through his infectious energy and breadth of academic enthusiasms and values. His outlook was shaped by the stimulus of a broad Scottish education that launched wide interests inside and outside science, including the history and literature of classical civilizations. His early research, with John Monteath Robertson FRS, was into structure determination by X-ray diffraction methods for single crystals, at a time when the full power of computers was just becoming realized for solution of the phase problem. With tenacity and originality, he then extended these approaches to materials that were to a greater or lesser extent disordered and even more difficult to solve because their diffraction patterns were poorer in information content. He brought many problems to definitive and detailed conclusion in a field that had been notable for solutions that were partial or vague, especially with oriented fibres of DNA and RNA but also various polysaccharides and synthetic polymers. His first approach was to use molecular model building in combination with difference Fourier analysis. This was followed later, and to even greater effect, by a computer refinement method that he developed himself and called linkedatom least-squares refinement. This has now been adopted as the standard approach by most serious centres of fibre diffraction analysis throughout the world. After the 10 years in which he consolidated his initial reputation at the Medical Research Council Biophysics Unit at King's College, London, in association with Maurice Wilkins FRS, he moved to Purdue University in the USA, first as Professor of Biology then becoming successively Head of the Department of Biological Sciences and Vice-President for Research and Dean of the Graduate School. As well as continuing his research, he contributed to the transformation of biological sciences at that university and to the development of the university's general management. He finally returned to his roots in Scotland as Principal and Vice-Chancellor of the University of St Andrews, to draw on his now formidable experience of international scholarship and institutional management, to reshape the patterns of academic life and mission to sit more happily and successfully within an environment that had become beset with conflict and change. He achieved this without disturbance to the harmony and wisdom embodied in the venerable traditions of that ancient Scottish yet cosmopolitan university.


2021 ◽  
Vol 13 (10) ◽  
pp. 1903
Author(s):  
Zhihui Li ◽  
Jiaxin Liu ◽  
Yang Yang ◽  
Jing Zhang

Objects in satellite remote sensing image sequences often have large deformations, and the stereo matching of this kind of image is so difficult that the matching rate generally drops. A disparity refinement method is needed to correct and fill the disparity. A method for disparity refinement based on the results of plane segmentation is proposed in this paper. The plane segmentation algorithm includes two steps: Initial segmentation based on mean-shift and alpha-expansion-based energy minimization. According to the results of plane segmentation and fitting, the disparity is refined by filling missed matching regions and removing outliers. The experimental results showed that the proposed plane segmentation method could not only accurately fit the plane in the presence of noise but also approximate the surface by plane combination. After the proposed plane segmentation method was applied to the disparity refinement of remote sensing images, many missed matches were filled, and the elevation errors were reduced. This proved that the proposed algorithm was effective. For difficult evaluations resulting from significant variations in remote sensing images of different satellites, the edge matching rate and the edge matching map are proposed as new stereo matching evaluation and analysis tools. Experiment results showed that they were easy to use, intuitive, and effective.


2021 ◽  
Vol 133 ◽  
pp. 105945
Author(s):  
Zhuyun Li ◽  
Xin Ren ◽  
Xin Wang ◽  
Wanli You ◽  
Meilian Zhong ◽  
...  

Author(s):  
Eko Hadi Sujiono ◽  
Vicran Zharvan ◽  
Sultra Ade Poetra ◽  
Muthmainnah Muchtar ◽  
Abdi Manab Idris ◽  
...  

2020 ◽  
pp. 1-11
Author(s):  
Guo Yunfeng ◽  
Li Jing

In order to improve the effect of the teaching method evaluation model, based on the grid model, this paper constructs an artificial intelligence model based on the grid model. Moreover, this paper proposes a hexahedral grid structure simplification method based on weighted sorting, which comprehensively sorts the elimination order of candidate base complexes in the grid with three sets of sorting items of width, deformation and price improvement. At the same time, for the elimination order of basic complex strings, this paper also proposes a corresponding priority sorting algorithm. In addition, this paper proposes a smoothing regularization method based on the local parameterization method of the improved SLIM algorithm, which uses the regularized unit as the reference unit in the local mapping in the SLIM algorithm. Furthermore, this paper proposes an adaptive refinement method that maintains the uniformity of the grid and reduces the surface error, which can better slow down the occurrence of geometric constraints caused by insufficient number of elements in the process of grid simplification. Finally, this paper designs experiments to study the performance of the model. The research results show that the model constructed in this paper is effective.


1999 ◽  
Vol 55 (2) ◽  
pp. 188-196 ◽  
Author(s):  
R. Høier ◽  
C. R. Birkeland ◽  
R. Holmestad ◽  
K Marthinsen

Quantitative convergent-beam electron diffraction is used to determine structure factors and three-phase structure invariants. The refinements are based on centre-disc intensities only. An algorithm for parameter-sensitive pixel sampling of experimental intensities is implemented in the refinement procedure to increase sensitivity and computer speed. Typical three-beam effects are illustrated for the centrosymmetric case. The modified refinement method is applied to determine amplitudes and three-phase structure invariants in noncentrosymmetric InP. The accuracy of the results is shown to depend on the choice of the initial parameters in the refinement. Even unrealistic starting assumptions and incorrect temperature factor lead to stable results for the structure invariant. The examples show that the accuracy varies from 1 to 10° in the electron three-phase invariants determined and from 0.5 to 5% for the amplitudes. Individual phases could not be determined in the present case owing to spatial intensity correlations between phase-sensitive pixels. However, for the three-phase structure invariant, stable solutions were found.


Sign in / Sign up

Export Citation Format

Share Document