[Keywords]:computer application; Chinese information processing; speaker verification; feature normalization; score normalization; NIST speaker evaluation;
基于6个网页-相关网页
model score normalization 模型得分归一化
likelihood score normalization 似然得分归一化
unsupervised score normalization 非监督得分规整
In this paper,two compensation approaches based on feature normalization and score normalization are presented,respectively.
本文提出了从特征规整和评分规整两个方面进行补偿的方法。
参考来源 - 基于特征规整和评分规整的说话人确认研究 in CIn decision making part, score normalization was studied in mathematical and physical aspects. Because training data was limited, our normalization method was in testing stage (like T-norm), and only top 20% scores was selected to compute the normalization score.
认证系统得判决部分主要探究了得分归一化的理论意义和实现方案,尤其征对文本相关的认证系统数据量小的特点,采用了测试归一化的方法,同时选择前几名的得分作为归一化得分,并且在归一化计算时忽略得分的方差。
参考来源 - 文本相关的说话人认证系统·2,447,543篇论文数据,部分数据来源于NoteExpress
In this paper, two compensation approaches based on feature normalization and score normalization are presented, respectively.
本文提出了从特征规整和评分规整两个方面进行补偿的方法。
In large scale standardized tests, the test scores, before release, have to undergo a series of transformation processes, including score weighting, score equating, and score normalization, etc.
为了使分数带有大量信息,便于用户使用,大规模标准化考试在最终报道考试成绩前,分数要经过加权处理、等值处理、正态化转换等一系列转换过程。
应用推荐