摘要
微表情是一闪而逝的真实表情,在谈判过程中,双方对谈判条件的真实接受程度和内心底线往往不会表现出来,但这种刻意的压制和隐藏却会被微表情所暴露。本文首先一段谈判视频中截取一系列图像,结合改进的SSD算法和比率检验法,对经HSV色彩空间转换后,剔除了环境背景的微表情图片进行识别。通过模拟谈判情境设计与实践,利用计算机能准确快速地识别微表情,经过微表情训练的实验对象能达成更有利的谈判结果。
Micro-expressions are fleeting true expressions. During negotiations, the real acceptance and inner bottom lines of both sides are often hidden, but this is deliberately suppressed and hidden, and it will be exposed to micro-expressions. Firstly, get interception photos from negotiators' videos. Then after the HSV color space conversion and the exclusion of the environmental background of micro -expressions, using improved SSD algorithm, combined with ratio test method, to recognize the images. After designed and practiced the simulation of negotiation situations, using the computer can recognize micro-expressions accurately and rapidly, with the help of microexpression recognition training, we will achieve better results in negotiations.
作者
马银忠
孙君
陶红
沈玉婷
史辛柱
MA Yin-zhong;SUN Jun;TAO Hong;SHEN Yu-ting;SHI Xin-zhu(Nanjing Institute of Technology,Nanjing 211167,China)
出处
《价值工程》
2016年第33期186-188,共3页
Value Engineering
基金
2016南京工程学院大学生创新项目+微表情识别及其在谈判过程中的应用+TB201601053
2016南京工程学院大学生创新项目+快递分拣设备+TB201601055
2016江苏省高等学校大学生创新创业训练计划项目+校园快递分拣设备+201611276068X
2015江苏省高等学校大学生创新创业训练计划项目+智能自动张合蚊帐+201511276062
关键词
微表情
识别
SSD算法
色彩空间
micro-expressions
recognition
SSD algorithm
color space