营业厅刷脸业务终端语音交互识别效果评估  

Evaluation of Speech Interactive Recognition Effect of Face Brushing Service Terminal in Business Hall

作　　者：黄媚 HUANG Mei(Shenzhen Power Supply Bureau Co.,Ltd.,Shenzhen 518001 China)

机构地区：[1]深圳供电局有限公司,广东深圳518001

出　　处：《自动化技术与应用》2023年第12期172-176,共5页Techniques of Automation and Applications

摘　　要：为提升营业厅刷脸业务终端语音交互识别效果,提出一种基于Hash算法的营业厅刷脸业务终端语音交互识别效果评估方法。计算连续丢包的数量,识别业务终端语音交互特征。确定评估对象因素集、评语等级集,构建基于Hash算法的识别效果评估模型。根据各个衡量因素隶属度,获取综合评估结果,并划分评估等级。实验结果可知,该方法在剔除编码造成的语音损伤因子后,评估结果与标准结果存在较大程度相似性,说明使用该评估方法的评估效果较好。In order to improve the effect of voice interactive recognition of business hall face brushing service terminal,this paper proposes an evaluation method of voice interactive recognition effect of business hall face brushing service terminal based on Hash algo-rithm.It calculates the number of continuous packet losses and identify the voice interaction characteristics of the service termi-nal.The factor set and comment level set of the evaluation object are determined,and the recognition effect evaluation model based on hash algorithm is constructed.According to the membership of each measurement factor,the comprehensive evaluation results are obtained and the evaluation grades are divided.The experimental results show that after eliminating the speech damage factors caused by coding,the evaluation results are similar to the standard results,indicating that the evaluation effect of this eval-uation method is good.

关 键 词：HASH算法 营业厅刷脸 业务终端 语音交互 识别效果评估 

分 类 号：TP391.42[自动化与计算机技术—计算机应用技术]