项目名称: 仿人机器人人性化表情交互理论与实现
项目编号: No.61273348
项目类型: 面上项目
立项/批准年度: 2013
项目学科: 自动化技术、计算机技术
项目作者: 黄强
作者单位: 北京理工大学
项目金额: 78万元
中文摘要: 仿人机器人从外形、举止、表情等方面与人类极为相似,具有很好的亲和性。为了使仿人机器人得到更广泛的应用,进入人们的工作、生活,要求仿人机器人具有人类习惯的、自然的、具有情感的表情交互能力。解决仿人机器人表情建模、辨识、表达等基础理论和关键技术,是解决机器人表情交互的必要前提。目前国内外对于机器人表情的研究刚起步,但考虑心理特性的高仿真度仿人机器人表情设计理论与方法的研究还未开展,这是一个亟待解决的基础科学问题。 本项目研究和借鉴人的情感产生机理、过程,人类表情与仿人机器人表情之间的映射关系,提取人类表情特征,建立仿人机器人表情模型;建立表情识别和推断的新算法,提高识别的准确性和实时性;提出基于心情、情感以及个性等心理因素的面向仿人机器人的表情表达决策模型,实现具有心理特性的仿人机器人的表情表达;在已研制的高仿真度仿人机器人上进行实验验证,为机器人"高仿真"地与人自然交互提供基础理论和方法。
中文关键词: 仿人机器人;人类表情分析;表情识别;表情建模;
英文摘要: Humanoid robots are highly similar to human beings in appearance, behaviors, facial expressions. Compared with other forms of robots, they look more friendly. In order to expand the utilization of humanoid robots and let them play a key role in our daily lives, they should have the same communication capabilities as humans, for example, facial expression interaction. The primary issues we need to solve before achieving human-like facial expression interaction is how to model, recognize and realize human expressions. While there have been recent work on robot facial expression, researchers have not studied facial expressions considering emotion, which is an urgent issue to be solved. In this project, we study the generation mechanism of human's affection, and map the relationship of facial expression between human beings and humanoids. Then we will extract characteristics of human expressions and model it. Expression recognition will be used to increase accuracy and decrease delay. Expression showing decision model based on mood, emotion and personality will be proposed to achieve humanoid facial expressions, which feature psychologic characters. Finally, we will demonstrate our proposed methods on a humanoid robot.
英文关键词: Humanoids Robot;Facial Expression Analysis;Facial Expression Recognition;Facial Expression Model;