The system is based on continuous hidden Markov model. Two-pass feature extraction and beam search strategy are also employed to reduce the consumption of the RAM and enhance the recognition rate.
系统采用连续隐含马尔科夫(CHMM)算法,运用了分阶段提取特征、束搜索等策略,在保证系统识别性能的同时大大降低了内存消耗,提高了识别速度,识别率在98.5%以上,识别时间在0.5倍实时以下。
参考来源 - 语音识别SoC UniLite的系统设计 in C·2,447,543篇论文数据,部分数据来源于NoteExpress
以上来源于: WordNet
Beam search is a heuristic search algorithm that is an optimization of best-first search that reduces its memory requirement.
定向搜索是一种启发式搜索算法,它是对减少内存需求的最佳优先搜索的优化算法。
Based on phrase-based statistical machine translation, a dynamical programming beam search decoding algorithm is put forward combining multi futures model using log-liner model approach.
在基于短语的统计机器翻译的基础上,结合对数线性模型的思想加入多个特征模型,研究了一种动态规划的柱搜索解码算法。
Secondly, based on the analysis of the relationship between the processing tasks, a hybrid algorithm combined the filtered beam search algorithm with the local search algorithm is proposed.
其次,在对工序任务之间的逻辑关系深入分析的基础上,提出一种基于过滤束算法思想与基本邻域搜索算法相结合的混合算法。
应用推荐