首页 | 本学科首页   官方微博 | 高级检索  
相似文献
 共查询到10条相似文献,搜索用时 140 毫秒
1.
根据人脸的生理结构特点,使用分块造型技术进行脸部建模,提出了一个基于"脸型特征"和"五官特征"的参数化肌肉模型,通过模型上控制顶点的属性值来控制脸部形成不同的表情,并利用MAYA MEL语言实现了一个实时交互操作界面.这使得对脸部表情的研究更加方便、灵活,建模及其动画化也更加逼真、实用.  相似文献   

2.
In this paper, we formulate the problem of synthesizing facial animation from an input audio sequence as a dynamic audio-visual mapping. We propose that audio-visual mapping should be modeled with an input-output hidden Markov model, or IOHMM. An IOHMM is an HMM for which the output and transition probabilities are conditional on the input sequence. We train IOHMMs using the expectation-maximization(EM) algorithm with a novel architecture to explicitly model the relationship between transition probabilities and the input using neural networks. Given an input sequence, the output sequence is synthesized by the maximum likelihood estimation. Experimental results demonstrate that IOHMMs can generate natural and good-quality facial animation sequences from the input audio.  相似文献   

3.
Performance has a spontaneity and “aliveness” that can be difficult to capture in more methodical animation processes such as keyframing. Access to performance animation has traditionally been limited to either low degree of freedom characters or required expensive hardware. We present a performance-based animation system for humanoid characters that requires no special hardware, relying only on mouse and keyboard input. We deal with the problem of controlling such a high degree of freedom model with low degree of freedom input through the use of correlation maps which employ 2D mouse input to modify a set of expressively relevant character parameters. Control can be continuously varied by rapidly switching between these maps. We present flexible techniques for varying and combining these maps and a simple process for defining them. The tool is highly configurable, presenting suitable defaults for novices and supporting a high degree of customization and control for experts. Animation can be recorded on a single pass, or multiple layers can be used to increase detail. Results from a user study indicate that novices are able to produce reasonable animations within their first hour of using the system. We also show more complicated results for walking and a standing character that gestures and dances.  相似文献   

4.
该文主要介绍了一个支持多模即时控制的角色动画库,包括其整体架构、脚本格式和动画引擎。提出了一种通过以XML作为主要格式的角色定义,将角色、骨架和动作定义相分离以提高动画库的复用性的方法。采用动作有限状态机使得系统能够通过一系列的动画指令来控制动画角色的行为,简化了对动画角色的控制。同时系统通过支持多模输入并在虚拟化身上做出即时反应,增强了虚拟现实应用中的沉浸感、临场参与感和交互性。  相似文献   

5.
6.
We present a physics-based approach to generate 3D biped character animation that can react to dynamical environments in real time. Our approach utilizes an inverted pendulum model to online adjust the desired motion trajectory from the input motion capture data. This online adjustment produces a physically plausible motion trajectory adapted to dynamic environments, which is then used as the desired motion for the motion controllers to track in dynamics simulation. Rather than using Proportional-Derivative controllers whose parameters usually cannot be easily set, our motion tracking adopts a velocity-driven method which computes joint torques based on the desired joint angular velocities. Physically correct full-body motion of the 3D character is computed in dynamics simulation using the computed torques and dynamical model of the character. Our experiments demonstrate that tracking motion capture data with real-time response animation can be achieved easily. In addition, physically plausible motion style editing, automatic motion transition, and motion adaptation to different limb sizes can also be generated without difficulty.  相似文献   

7.
提出一种基于状态异步动态贝叶斯网络模型(SA-DBN)的语音驱动面部动画合成方法。提取音视频语音数据库中音频的感知线性预测特征和面部图像的主动外观模型(AAM)特征来训练模型参数,对于给定的输入语音,基于极大似然估计原理学习得到对应的最优AAM特征序列,并由此合成面部图像序列和面部动画。对合成面部动画的主观评测结果表明,与听视觉状态同步的DBN模型相比,通过限制听觉语音状态和视觉语音状态间的最大异步程度,SA-DBN可以得到清晰自然并且嘴部运动与输入语音高度一致的面部动画。  相似文献   

8.
Algorithm animation would seem to be a useful tool for teaching algorithms. However, previous empirical studies of using algorithm animation have produced mixed results. This paper presents an empirical study in which the subjects programmed the algorithm which they had seen animated. The results of the experiment indicate that combining the animation with the implementation of the algorithm was an effective way to teach the animation, and also produced transfer effects for general recursion problems.  相似文献   

9.
目的 动漫制作中线稿绘制与上色耗时费力,为此很多研究致力于动漫制作过程自动化。目前基于数据驱动的自动化研究工作快速发展,但并没有一个公开的线稿数据集可供使用。针对真实线稿图像数据获取困难,以及现有线稿提取方法效果失真等问题,提出基于循环生成对抗网络的线稿图像自动提取模型。方法 模型基于循环生成对抗网络结构,以解决非对称数据训练问题。然后将不同比例的输入图像及其边界图输入到掩码指导卷积单元,以自适应选择网络中间特征。同时为了进一步提升网络提取线稿的效果,提出边界一致性约束损失函数,确保生成结果与输入图像在梯度变化上的一致性。结果 在公开的动漫彩色图像数据集Danbooru2018上,应用本文模型提取的线稿图像相比于现有线稿提取方法,噪声少、线条清晰且接近真实漫画家绘制的线稿图像。实验中邀请30名年龄在2025岁的用户,对本文以及其他4种方法提取的线稿图像进行打分。最终在30组测试样例中,本文方法提取的线稿图像被认为最佳的样例占总样例84%。结论 通过在循环生成对抗网络中引入掩码指导单元,更加合理地提取彩色图像的线稿图像,并通过对已有方法提取效果进行用户打分证明,在动漫线稿图像提取中本文方法优于对比方法。此外,该模型不需要大量真实线稿图像训练数据,实验中仅采集1 000幅左右真实线稿图像。模型不仅为后续动漫绘制与上色研究提供数据支持,同时也为图像边缘提取方法提供了新的解决方案。  相似文献   

10.
提出了一种用于动态序列合成的统计模型-基于核密度估计的隐马尔可夫模型,给定一个输入动态序列,该模型可以自动产生被控的输出动态序列,文中提出的模型是一种以非参数化概率密度估计作为观测模型的隐马尔可夫模型,该模型对输入和受控输出序列的联合概率分布进行建模,并利用基于核函数的概率密度估计来学习联合概率分布的细节信息,文中详细地讨论了该模型的学习和合成算法,并利用该模型实现了一个虚拟指挥系统,即给定一段音乐,系统可以自动生成相关的乐队指挥动作,该文利用该系统对不同风格和节拍的音乐做了实验,实验结果验证了算法的有效性。  相似文献   

设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号