字典神经网络方法快速湍流烟雾合成

目的 基于物理的烟雾模拟是计算机图形学的重要组成部分,渲染具有细小结构的高分辨率烟雾,需要大量的计算资源和高精度的数值求解方法。针对目前高精度湍流烟雾模拟速度慢,仿真困难的现状,提出了基于字典神经网络的方法,能够快速合成湍流烟雾,使得合成的结果增加细节的同时,保持高分辨率烟雾结果的重要结构信息。方法 使用高精度的数值仿真求解方法获得高分辨率和低分辨率的湍流烟雾数据,通过采集速度场局部块及相应的空间位置信息和时间特征生成数据集, 设计字典神经网络的网络架构,训练烟雾高频成分字典预测器,在GPU（graphic processing unit）上实现并行化,快速合成高分辨率的湍流烟雾结果。结果 实验表明,基于字典神经网络的方法能够在非常低分辨率的烟雾数据下合成空间和时间上连续的高分辨率湍流烟雾结果,效率比通过在GPU平台上直接仿真得到高分辨率湍流烟雾的结果快了一个数量级,且合成的烟雾结果与数值仿真方法得到的高分辨率湍流烟雾结果足够接近。结论 本文方法解决了烟雾的上采样问题,能够从非常低分辨率的烟雾仿真结果,通过设计基于字典神经网络结构以及特征描述符编码烟雾速度场的局部和全局信息,快速合成高分辨率湍流烟雾结果,且保持高精度烟雾的细节,与数值仿真方法的对比表明了本文方法的有效性。     

Fast and Scalable CPU/GPU Collision Detection for Rigid and Deformable Surfaces

We present a new hybrid CPU/GPU collision detection technique for rigid and deformable objects based on spatial subdivision. Our approach efficiently exploits the massive computational capabilities of modern CPUs and GPUs commonly found in off‐the‐shelf computer systems. The algorithm is specifically tailored to be highly scalable on both the CPU and the GPU sides. We can compute discrete and continuous external and self‐collisions of non‐penetrating rigid and deformable objects consisting of many tens of thousands of triangles in a few milliseconds on a modern PC. Our approach is orders of magnitude faster than earlier CPU‐based approaches and up to twice as fast as the most recent GPU‐based techniques.     

基于物理的流体模拟动画综述

基于物理的流体模拟近年来成为计算机动画领域的一个研究热点.回顾了该领域中基于物理的流体模拟的发展情况,总结了该研究方向所采用的各类方法,并结合各种现象的特点分门别类地详细展开.其方法总体上可以分为欧拉法和拉格朗日法,涉及的现象包括烟雾、火焰、爆炸、波浪、气泡以及自由运动界面等.最后展望了未来发展的3个重点：细节策略、加速策略和控制策略,以使整个模拟能够更好地满足人们对真实感、实时性以及灵活性的需求.     

有机化学综合设计型实验的计算机模拟

利用Flash动画和Director多媒体技术,开发制作了有机化学实验计算机模拟软件。在系统模拟有机化学实验的基本知识和单元实验的基础上,重点对一些综合型、设计型实验进行了仿真模拟,内容包括12项单元实验、42项综合实验和2项示范性设计实验。     

用于动态序列合成的运动纹理模型

提出了一种新的用于动态序列合成的分层统计模型——运动纹理模型．该模型使用统计方法自动分析动态序列,可以合成与原始样本数据统计特性相同的新的动态序列．运动纹理模型是一个两层模型．其中包含的两层分别为基元层和基元分布层．模型中用线性动态系统来表示单个基元．而对基元的统计分布则由相关转移矩阵来描述．该文详细地讨论了如何通过最大似然准则来学习运动纹理模型的方法,并描述了如何用运动纹理模型自动合成动态序列的过程．该文通过使用运动纹理模型合成舞蹈动作和视频序列的一些实验验证了该模型的有效性．     

Fostering elementary school students' understanding of simple electricity by combining simulation and laboratory activities

Abstract   Computer simulations and laboratory activities have been traditionally treated as substitute or competing methods in science teaching. The aim of this experimental study was to investigate if it would be more beneficial to combine simulation and laboratory activities than to use them separately in teaching the concepts of simple electricity. Based on their pre-test performances, 66 elementary school students were placed into three different learning environments: computer simulation, laboratory exercise and a simulation–laboratory combination. The results showed that the simulation–laboratory combination environment led to statistically greater learning gains than the use of either simulation or laboratory activities alone, and it also promoted students' conceptual understanding most efficiently. There were no statistical differences between simulation and laboratory environments. The results highlight the benefits of using simulation along with hands-on laboratory activities to promote students' understanding of electricity. A simulation can help students to first understand the theoretical principles of electricity; however, in order to promote conceptual change, it is necessary to challenge further students' intuitive conceptions by demonstrating through testing that the laws and principles that are discovered through a simulation also apply in reality.     

计算机生成3维卡通烟尘动画

为了能够表现不同透视和视点下的3维卡通烟尘动画,给出了一个计算机自动生成3维卡通烟尘动画的模型.该模型首先对手工绘制的卡通烟尘进行分类,然后通过构建它们的骨架来定义出各种类型卡通烟尘的静态结构构与动态结构,最后在骨架的基础上,通过简单的模型绘制出卡通烟尘的形态.该模型的特点是把骨架定义在3维空间上,这不仅仅可以表现不同透视和视点下的各种卡通烟尘效果,而且避免了手工绘制相应动画效果时的重复劳动.     

Animation modeling with petri nets

This paper introduces the use of Petri Nets as a modeling and analysis tool for animation environments. Firstly, the original formulation for Petri Nets is applied in two animation situations, one modeled as a state machine and another exploring interdependent transitions. Increasing the complexity level, some modeling extensions are discussed and more sophisticated animation examples are studied.     

A behavioural test-bed using a dataglove input device

In many cases, traditional hand-drawn animation has been replaced by computer technology. Computer-supported approaches can essentially be characterised as two interface types: keyframing and coding. However, these two interface types offer limited editing ability for scene animation applications, which usually consist of a large testing space of similar behaviours. The testing cycle, using either predefined keyframe sequences or general coding interface, tends to be costly and time consuming. This paper reports work which uses the DataGlove device to support and test variable schooling behaviours of fish in a virtual marine world. This is put forward as a representative example of scene animation. The glove-based interface places the user as a participant in the behavioural simulation process. In the work, hand shapes and motions are recognised and used for either event triggering or role switching. The specific shapes and motions of the user's hand trigger control signals or commands through a menu-based interface. The hand can itself be used to simulate an object in the scene. The object, which can be either static or dynamic, participates in the control process. Using hand movements in this way allows the user to interactively specify the paths of moving objects in the scene, and also creates a diversity of dynamic situations which can be useful for testing variable scene behaviours. The application presented in this paper looks at examples of controlling fish behaviour in a limited pond environment controlled by glove-based interaction.     

流体动画方法综述

介绍了二维流体和三维流体动画方法。流体动画制作分为两步：流体表面建模和流体绘制。二维流体建模主要采用参数建模方法,基于傅立叶变换原理,合成波动的二维流体表面。三维流体建模一般基于Navier-Stokes流体方程,利用流体内部的速度建立流体表面的方程。流体绘制主要采用光线跟踪方法。流体动画的挑战和发展方向在于：流体动画控制、实时绘制、细节表现。     

基于IMEX的织物动态仿真的近似解法

织物在空间运动的刚性特征始终是困扰织物动态仿真的难题．显式方法简单灵活,易于实现,但受稳定因素影响,无法实现具有刚性特征的织物动态模拟;隐式方法稳定性好,却忽略了非线性因素,而且计算复杂,直接影响到仿真的最终结果和实际效率．针对这一问题,提出了基于隐式一显式的近似解法,该方案从系统受力形变的非线性特征出发,将质点受力分为线性和非线性两部分,线性部分采用隐式解法,非线性部分利用显式解法,线性方程组的求解则运用近似解法．实验结果表明,该方法兼具两种方法的优点,既保留了隐式方法的稳定性,又充分利用了显式方法的简易性处理非线性特征,从而从真正意义上解决织物仿真中的刚性问题．     

结合一种面-面碰撞检测算法的服装动态模拟

实时的服装动态模拟一直是计算机动画的研究热点。在回顾弹性模型及其碰撞问题的相关研究工作的基础上,采用基于质点－弹簧模型的动态模拟方法,产生了虚拟3D模特表面的服装动态效果。其中考虑了织物的非理想弹性属性和变化的空气流作用力,并针对系统实现的瓶颈－服装和人体的碰撞问题,提出一种碰撞检测算法。