algorithm translation
1.Research and Implementation of A~* Search Algorithm for Chinese-English Statistical Machine Translation
2.Powell algorithm, which does not compute gradients, and Brent algorithm were used to search the optimal rotation and translation parameters.
4.Powell algorithm which does not compute gradients and Brent algorithm were used to search the optimal rotation and translation parameters, while multi-resolution searching strategy leads to a fast image registration.
5.In this paper,a mixed language model integrating rule-based grammer and statistical language model for translating Braille to Chinese is introduced,this technology integrating Markov language model based on statistic and quantized rules,employing Viterbi algorithm to find the best Chinese string,and the result of the translation is good.
6.This paper is intended to focus on the discussion of the storage structure and search algorithm of an integrated lexicon to be used in our Matlink English to Chinese Machine Translation System which is now under development.
7.Based on analyzing the 64-bit address space of the IA64 architecture, this paper analyzes the application requirements of the Single Address Space(SAS) and Multiple Address Space(MAS) models, studies the two address mapping mechanisms based on long-format and short-format VHPT, and realizes a hash address algorithm of the two formats of VHPT. This enhances the virtual address translation performance.
8.the protocol translation implements the algorithm of IP header between IPv4 and IPv6,between ICMPv4 and ICMPv6 and the upate of checksum field of TCP/UDP. In the end,We implemented translation gateway by VC++6.0 and Windows DDK,meanwhile, examined Latency of traslating packet,Bandwidth of network and the operation of familiar application.
9.3. Introduce and implement the static algorithm for jump table recovery for IA-64. The algorithm is proved by ITA system to be correct and effective, because it can resolve the resumption for jump tables and target address in various compilers for IA-64, and improve the code coverage rate of static translation.
10.Experiment results show that the translation registration errors(+0.254,+0.083)are far less than one pixel, and the stabilizing time for single frame image on some definite hardware platform only needs14.7ms, just 1/3 of some other algorithm with same precision, which fully satisfies the real-time requirement and the processing precision of imaging system on vehicle.
