节点文献
一种评价噪声烦恼度的新方法
A novel appraisal strategy of noise annoyance
【作者】 时荣伟; 闫靓; 周金丽; 李斐远; 李俊; 李志远;
【Author】 SHI Rongwei;YAN Liang;ZHOU Jingli;LI Feiyuan;LI Jun;LI Zhiyuan;School of Marine Science and Technology, Northwestern Polytechnical University;Guangzhou Qijing Environmental Technology Co.Ltd;Anhui Weiwei Vibration and Noise Reduction Technology Research Institute;
【机构】 西北工业大学航海学院; 广州启境环保科技有限公司; 安徽微威减震降噪技术研究院;
【摘要】 本文围绕噪声烦恼度主观测量方法的应用短板,提出"以听音者面部表情的泛化特征构建普适性的面部表情常模,表征噪声引发的烦恼情绪;以听音者面部表情的变化评估噪声烦恼度"的新方法。论文首先研究了与复合负性(消极)情绪——烦恼——有关的3类基本负性情绪(愤怒、厌恶、悲伤)的面部表情的共性特征,参考面部动作编码系统(Facial Action Coding System, FACS),对其中的关键性面部动作单元(Action Unit, AU)进行识别、提取与重塑,确定了能够量化负性情绪人脸面部表情的24维特征,人工绘制出具有复合情绪特点的负性情绪面部表情常模。以此为基础,通过改变24维面部特征的具体数值,构建了可用于度量五种情绪程度水平的面部表情量表。最后,借助大规模网络调查对相关理论研究成果的有效性进行了验证。结果表明,相比于传统评估方法,以文中设计开发的面部表情常模与面部表情尺度进行噪声烦恼度的具像化表征与量化评估,具有实时灵敏、高效直观、绿色经济的显著优势。
【Abstract】 Based on the short-board of noise annoyance appraisal procedure, a new method is proposed for noise annoyance appraisal which including an universal facial expression norm with the generalized feature of the listener’s facial expression to represent the annoyance emotion caused by noise, and its corresponding facial expression scale. Firstly, the common features of three basic negative emotions(anger/disgust/sadness) related to the complex negative emotion were studied by facial expression feature extraction. Referring to the facial action coding system(Facial Action Coding System, FACS), the key facial action units(Action Unit,AU) in the negative facial expression norm were identified and extracted. Then the corresponding facial expression scale with five levels for emotion assessment by changing the specific values of 24-dimensional facial features was established. Finally, with the help of large-scale network investigation, the validity of relevant theoretical research results is verified. Compared with the traditional evaluation method, the visual representation and quantitative evaluation of noise annoyance by facial expression norm and facial expression scale has the advantages of real-time sensitivity, high efficiency and green economy.
【Key words】 noise annoyance; facial expression; facial action coding system; facial expression scale;
- 【会议录名称】 2020’中国西部声学学术交流会论文集
- 【会议名称】2020’中国西部声学学术交流会
- 【会议时间】2020-08-20
- 【会议地点】中国甘肃酒泉
- 【分类号】TP391.41;B845.6
- 【主办单位】四川省声学学会、上海声学学会、山东声学学会、黑龙江省声学学会、重庆声学学会、西安声学学会、陕西省声学学会、全国声学标准化技术委员会、中国声学学会环境声学分会